app/models/scoring.py

@@ -55,13 +55,14 @@ class ATSScore(StrictModel):
     matched_keywords: list[str] = Field(default_factory=list, max_length=30)
     missing_keywords: list[str] = Field(default_factory=list, max_length=30)
     summary_critique: str = Field(min_length=1, max_length=500)
+    professional_summary: str | None = Field(default=None, max_length=500)
 
     @field_validator("matched_keywords", "missing_keywords", mode="before")
     @classmethod
     def _normalize(cls, value: Any) -> Any:
         return _normalize_keywords(value)
 
-    @field_validator("candidate_name", "job_title", "current_company", mode="before")
+    @field_validator("candidate_name", "job_title", "current_company", "professional_summary", mode="before")
     @classmethod
     def _blank_profile_text_to_none(cls, value: Any) -> Any:
         if isinstance(value, str):

app/prompts/ats.py

@@ -52,6 +52,10 @@ entries; null if neither is stated.
 (for example "6 years of experience"), use that stated number; otherwise compute \
 whole years only from dates or durations explicitly stated in the resume; null \
 whenever neither is available.
+- professional_summary: one or two sentences naming the candidate's tech-stack \
+speciality and functional department from the resume alone. Ignore the job \
+description. This is not summary_critique. Null if the resume does not evidence \
+either a stack or a department.
 
 Return concise, evidence-based fields matching the supplied JSON schema. \
 matched_keywords must contain only skills that appear in the resume, written with the \

backend/credentials/application_default_credentials.json

@@ -2,10 +2,10 @@
   "type": "authorized_user",
   "client_id": "679334897177-ufal3rogbg8cgm3qcren6pv20phqd7nl.apps.googleusercontent.com",
   "client_secret": "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",
-  "refresh_token": "1//03C8LMkk-9mSyCgYIARAAGAMSNwF-L9IrAcPhShsp3cDeprSlqI-P6lArpmbyzu-PcKiLfQ5gC3H_MEk930IaKfewy3cxP3T0Oo8",
+  "refresh_token": "1//03Tu7LRVp_zBACgYIARAAGAMSNwF-L9IrIXwqqdEZGxpxULAKbtgCygCih8DHmO-ELciPtMV8VCVNyZCrO9l6veq0WPwT6fbcAC8",
   "universe_domain": "googleapis.com",
   "account": "ahmed.mujtaba@utopiabrands.com",
-  "token": "ya29.a0AdMD6EgKB22VSy--W0qCtRkMOYECCDhvL4c14xNUSvizbooOBC-ctQeyWR_XybUqa6PZvQ0csVqrIDR6e_uQazKuTAxKPLgpgbTmJ96-sHWbj_981xNrcWe6JsxIpQuGLX9GiKSGa8y5t50ZgWbDy0ECUoOQDvzlUq-hgNdLve1ECxDG4twpL3-2ZpGgskHlhuRL4-PQaCgYKARISARASFQHGX2MiOWOjgYvNnX5ZWhBoYqO9Cg0207",
-  "expiry": "2026-09-02T16:27:05Z",
+  "token": "ya29.a0AdMD6Eh9Kvd-ACJZT90CDywa396Zsrf84OWg17u8X-AVffmKhB0nuql60ail5cAY8XlkRuySHSZRKSdXQ7W3IM2dticjCoYgeMmVErMi5UAawUQAd6q0CEsCbi7EPnLTgraOXTAO1MRlWaHwU-R179t0GsAQwnlj9SWBC5Zgfu7Ubf8dGyBcuzg9zknfCF2zbmZFZOsaCgYKAV0SARASFQHGX2MiOeWelgu56Tql44hrVoy1iw0206",
+  "expiry": "2026-09-08T12:10:05Z",
   "quota_project_id": "hrms-ats-portal"
 }

backend/g_sheet/enums.py

@@ -207,6 +207,7 @@ class FormDataColumn(str, Enum):
     AREA_OF_RESIDENCE = "area_of_residence"
     RESIDING_CITY = "residing_city"
     CITY = "city"
+    PROFESSIONAL_SUMMARY = "professional_summary"
     RESIDING_COUNTRY = "residing_country"
     COMMUNICATION_SKILLS = "communication_skills"
     PREFERRED_TIMINGS = "preferred_timings"

backend/g_sheet/models.py

@@ -86,6 +86,7 @@ class FormData(SQLModel, table=True):
     residing_city: str | None = Field(default=None)
     residing_country: str | None = Field(default=None)
     city: str | None = Field(default=None)
+    professional_summary: str | None = Field(default=None)
     reapplied: list[str] = Field(default_factory=list, sa_type=JSONB, sa_column_kwargs={"server_default": "[]"})
     communication_skills: int | None = Field(default=None)
     preferred_timings: str | None = Field(default=None)
@@ -229,6 +230,17 @@ class FormData(SQLModel, table=True):
         result = await session.execute(select(cls).where(cls.id == rid))
         return result.scalars().first()
 
+    @classmethod
+    async def set_professional_summary(cls, session: AsyncSession, record_id, summary):
+        row = await cls.get_form_data_by_id(session, record_id)
+        if not row:
+            return None
+        row.professional_summary = (summary or "").strip() or None
+        session.add(row)
+        await session.commit()
+        await session.refresh(row)
+        return row
+
     @classmethod
     async def get_with_job(cls, session: AsyncSession, record_id, job_post_id):
         """Form row + one job it may be scored against (suggested or assigned)."""
@@ -390,6 +402,21 @@ class FormData(SQLModel, table=True):
         result = await session.execute(statement)
         return result.scalars().all()
 
+    @classmethod
+    async def list_on_hold_scan_rows(cls, session: AsyncSession, sheet=None):
+        """On-Hold Sheet Forms: id + email. Entire catalogue, optional sheet tab."""
+        statement = select(cls.id, cls.candidate_email)
+        for clause in cls._filters(sheet=sheet, no_suggestions=True):
+            statement = statement.where(clause)
+        result = await session.execute(statement)
+        rows = []
+        for record_id, email in result.all():
+            rows.append({
+                "id": record_id,
+                "email": (email or "").strip().lower() or None,
+            })
+        return rows
+
     @classmethod
     async def list_by_emails(cls, session: AsyncSession, emails):
         """Sheet applicants for these addresses. Promoted rows are omitted —

backend/g_sheet/scoring.py

@@ -17,7 +17,7 @@ from app.services.pdf import ExtractedResume
 from app.services.scoring import score_batch
 from db_setup import session_scope
 from g_sheet.models import FormData
-from inbox.models import AtsResults
+from inbox.models import AtsResults, InboxRescanRun
 from job.candidate.plugins import build_job_description, get_scorer, get_scoring_settings
 from job.job_post.models import JobPosts
 
@@ -46,6 +46,7 @@ def serialize_form_ats(row) -> dict:
         "job_post_id": str(row.job_post_id) if row.job_post_id else None,
         "overall_score": row.overall_score,
         "band": row.band or None,
+        "professional_summary": row.professional_summary or None,
         "computed_at": row.computed_at.isoformat() if row.computed_at else None,
     }
 
@@ -96,7 +97,7 @@ async def enqueue_form_row_scores(form_row) -> None:
     await enqueue_form_scores(form_row.id,job_ids)
 
 
-async def score_form_against_job(form_data_id: str, job_id: str) -> dict:
+async def score_form_against_job(form_data_id: str, job_id: str, rescan_run_id=None) -> dict:
     """Score one Sheet Forms CV against one job. Idempotent per (form, job)."""
     try:
         uuid.UUID(str(form_data_id))
@@ -143,6 +144,14 @@ async def score_form_against_job(form_data_id: str, job_id: str) -> dict:
         logger.warning("form ats failed form_data=%s job=%s code=%s", form_data_id, job_id, error)
         return {"status": "failed", "error_code": error}
 
+    summary = (result.professional_summary or "").strip() or None
+    run_id = None
+    if rescan_run_id not in (None, ""):
+        try:
+            run_id = uuid.UUID(str(rescan_run_id))
+        except (TypeError, ValueError):
+            run_id = None
+
     async with session_scope() as session:
         existing = await AtsResults.get_for_form_job(session, form_pk, job_pk)
         if existing is not None:
@@ -150,6 +159,7 @@ async def score_form_against_job(form_data_id: str, job_id: str) -> dict:
         job = await JobPosts.get_job_post_by_id(session, job_pk)
         if job is None or job.is_deleted:
             return {"status": "skipped", "reason": "job_gone"}
+        await FormData.set_professional_summary(session, form_pk, summary)
         await AtsResults.insert_result(session, {
             "inbox_id": None,
             "user_id": None,
@@ -159,8 +169,17 @@ async def score_form_against_job(form_data_id: str, job_id: str) -> dict:
             "overall_score": float(result.match_score),
             "band": _band(result.match_score),
             "model_name": settings.openai_model,
+            "professional_summary": summary,
+            "rescan_run_id": run_id,
             "is_current": True,
         })
+        if run_id:
+            await InboxRescanRun.append_summary(session, run_id, {
+                "kind": "form",
+                "record_id": str(form_pk),
+                "job_post_id": str(job_pk),
+                "professional_summary": summary,
+            })
     return {
         "status": "scored",
         "overall_score": result.match_score,

backend/inbox/app.py

@@ -74,6 +74,11 @@ class DuplicateBody(BaseModel):
     is_duplicate: bool
 
 
+class OnHoldRescanBody(BaseModel):
+    channel: str = "all"
+    sheet: str | None = None
+
+
 class ReadBody(BaseModel):
     read: bool = True
 
@@ -513,3 +518,38 @@ async def reply_email(
         raise
     except Exception as e:
         raise HTTPException(status_code=500,detail=str(e))
+
+
+@router.post("/inbox/rescan-on-hold")
+async def start_on_hold_rescan(
+    payload: OnHoldRescanBody,
+    current_user: dict = Depends(require_permission(PermissionTag.INBOX_EDIT)),
+    session: AsyncSession = Depends(get_session),
+):
+    """Score every On-Hold CV against all job posts. Poll GET until completed."""
+    try:
+        service=Email(session=session)
+        data=await service.start_on_hold_rescan(
+            channel=payload.channel,sheet=payload.sheet,current_user=current_user,
+        )
+        return JSONResponse(content={"data":data,"total":1,"status_code":200})
+    except HTTPException:
+        raise
+    except Exception as e:
+        raise HTTPException(status_code=500,detail=str(e))
+
+
+@router.get("/inbox/rescan-on-hold")
+async def fetch_on_hold_rescan(
+    run_id: str | None = Query(None),
+    current_user: dict = Depends(require_permission(PermissionTag.INBOX_VIEW)),
+    session: AsyncSession = Depends(get_session),
+):
+    try:
+        service=Email(session=session)
+        data=await service.get_on_hold_rescan(run_id=run_id)
+        return JSONResponse(content={"data":data,"total":1,"status_code":200})
+    except HTTPException:
+        raise
+    except Exception as e:
+        raise HTTPException(status_code=500,detail=str(e))

backend/inbox/models.py

@@ -682,6 +682,8 @@ class Inbox_Messages(SQLModel, table=True):
     source_channel_id: int | None = Field(default=None, foreign_key="source_channels.id")
     processing_state: str = Field(default="unread", sa_column_kwargs={"server_default": "unread"})
     city: str | None = Field(default=None)
+    # Job-agnostic tech-stack + department, overwritten on every completed ATS run.
+    professional_summary: str | None = Field(default=None)
     # Ingestion time — list screens order by this (newest first). server_default
     # backfills existing rows on the ALTER so NOT NULL is safe on a populated table.
     created_at: datetime = Field(
@@ -1137,6 +1139,23 @@ class Inbox_Messages(SQLModel, table=True):
         result = await session.execute(statement)
         return result.scalars().all()
 
+    @classmethod
+    async def list_on_hold_scan_rows(cls, session: AsyncSession):
+        """On-Hold email applications: id + sender + CV path. No TOAST columns."""
+        statement = cls._apply_filters(
+            select(cls.id, cls.message_from, cls.file_path),
+            no_suggestions=True,
+        )
+        result = await session.execute(statement)
+        rows = []
+        for record_id, email, file_path in result.all():
+            rows.append({
+                "id": record_id,
+                "email": (email or "").strip().lower() or None,
+                "file_path": (file_path or "").strip() or None,
+            })
+        return rows
+
     @classmethod
     async def get_inbox_message_by_id(cls, session: AsyncSession, record_id: str):
         try:
@@ -1158,6 +1177,18 @@ class Inbox_Messages(SQLModel, table=True):
         await session.refresh(row)
         return row
 
+    @classmethod
+    async def set_professional_summary(cls, session: AsyncSession, record_id, summary):
+        row = await cls.get_inbox_message_by_id(session, record_id)
+        if not row:
+            return None
+        text=(summary or "").strip() or None
+        row.professional_summary = text
+        session.add(row)
+        await session.commit()
+        await session.refresh(row)
+        return row
+
     @classmethod
     async def set_assigned_job_post(cls, session: AsyncSession, record_id, job_post_id):
         """Set or clear assigned_job_post_id; returns the row or None if missing."""
@@ -1813,6 +1844,9 @@ class AtsResults(SQLModel, table=True):
     job_post_id: uuid.UUID | None = Field(default=None, foreign_key="job_posts.id")
     # Sheet Forms scores: no inbox/user. Identity is (form_data_id, job_post_id).
     form_data_id: uuid.UUID | None = Field(default=None, index=True, foreign_key="form_data.id")
+    # On-Hold ReScan run that produced this row. NULL for assign-job / upload scores.
+    rescan_run_id: uuid.UUID | None = Field(default=None, index=True, foreign_key="inbox_rescan_runs.id")
+    professional_summary: str | None = Field(default=None)
     overall_score: float = Field(default=0.0)
     band: str = Field(default="")
     is_current: bool = Field(default=True)
@@ -1961,6 +1995,96 @@ class AtsResults(SQLModel, table=True):
             grouped.setdefault(row.form_data_id, []).append(row)
         return grouped
 
+    @classmethod
+    async def job_ids_by_emails(cls, session: AsyncSession, emails) -> dict:
+        """{email: {job_post_id, ...}} for any prior ATS score of these people."""
+        from g_sheet.models import FormData
+
+        lowers = sorted({(e or "").strip().lower() for e in (emails or []) if (e or "").strip()})
+        if not lowers:
+            return {}
+        grouped: dict[str, set[str]] = {}
+
+        def _add(email, job_id):
+            key = (email or "").strip().lower()
+            if not key or job_id is None:
+                return
+            grouped.setdefault(key, set()).add(str(job_id))
+
+        inbox_rows = await session.execute(
+            select(func.lower(Inbox_Messages.message_from), cls.job_post_id)
+            .join(Inbox, cls.inbox_id == Inbox.id)
+            .join(Inbox_Messages, Inbox.message_id == Inbox_Messages.id)
+            .where(func.lower(Inbox_Messages.message_from).in_(lowers))
+            .where(cls.job_post_id.is_not(None))
+        )
+        for email, job_id in inbox_rows.all():
+            _add(email, job_id)
+
+        form_rows = await session.execute(
+            select(func.lower(FormData.candidate_email), cls.job_post_id)
+            .join(FormData, cls.form_data_id == FormData.id)
+            .where(func.lower(FormData.candidate_email).in_(lowers))
+            .where(cls.job_post_id.is_not(None))
+        )
+        for email, job_id in form_rows.all():
+            _add(email, job_id)
+
+        user_rows = await session.execute(
+            select(func.lower(Users.email), cls.job_post_id)
+            .join(Users, cls.user_id == Users.id)
+            .where(func.lower(Users.email).in_(lowers))
+            .where(cls.job_post_id.is_not(None))
+        )
+        for email, job_id in user_rows.all():
+            _add(email, job_id)
+        return grouped
+
+    @classmethod
+    async def job_ids_for_messages(cls, session: AsyncSession, message_ids) -> dict:
+        """{message_id: {job_post_id, ...}} scored for these inbox_messages rows."""
+        keys = []
+        for raw in message_ids or []:
+            uid = cls._as_uuid(raw)
+            if uid is not None:
+                keys.append(uid)
+        if not keys:
+            return {}
+        result = await session.execute(
+            select(Inbox.message_id, cls.job_post_id)
+            .join(Inbox, cls.inbox_id == Inbox.id)
+            .where(Inbox.message_id.in_(keys))
+            .where(cls.job_post_id.is_not(None))
+        )
+        grouped: dict[str, set[str]] = {}
+        for mid, job_id in result.all():
+            if mid is None or job_id is None:
+                continue
+            grouped.setdefault(str(mid), set()).add(str(job_id))
+        return grouped
+
+    @classmethod
+    async def job_ids_for_forms(cls, session: AsyncSession, form_ids) -> dict:
+        """{form_data_id: {job_post_id, ...}} scored for these Sheet Forms rows."""
+        keys = []
+        for raw in form_ids or []:
+            uid = cls._as_uuid(raw)
+            if uid is not None:
+                keys.append(uid)
+        if not keys:
+            return {}
+        result = await session.execute(
+            select(cls.form_data_id, cls.job_post_id)
+            .where(cls.form_data_id.in_(keys))
+            .where(cls.job_post_id.is_not(None))
+        )
+        grouped: dict[str, set[str]] = {}
+        for fid, job_id in result.all():
+            if fid is None or job_id is None:
+                continue
+            grouped.setdefault(str(fid), set()).add(str(job_id))
+        return grouped
+
     @classmethod
     async def resolve_identity(cls, session: AsyncSession, email, candidate_id):
         """XOR identity for a score row from the scored candidate's email.
@@ -2046,6 +2170,16 @@ class AtsResults(SQLModel, table=True):
         await session.commit()
         return row
 
+    @classmethod
+    async def list_by_rescan_run(cls, session: AsyncSession, rescan_run_id):
+        uid = cls._as_uuid(rescan_run_id)
+        if uid is None:
+            return []
+        result = await session.execute(
+            select(cls).where(cls.rescan_run_id == uid).order_by(cls.computed_at.desc())
+        )
+        return list(result.scalars().all())
+
 
 class MailboxSyncRun(SQLModel, table=True):
     """One Outlook mailbox sync job — survives tab close because work runs in Taskiq.
@@ -2125,3 +2259,96 @@ class MailboxSyncRun(SQLModel, table=True):
             await session.commit()
             await session.refresh(row)
         return row
+
+
+class InboxRescanRun(SQLModel, table=True):
+    """On-Hold catalogue ATS rescan — one row the Inbox ReScan button polls."""
+
+    __tablename__ = "inbox_rescan_runs"
+
+    id: uuid.UUID = Field(default_factory=uuid.uuid4, primary_key=True)
+    status: str = Field(default="queued", index=True)  # queued|running|scoring|completed|failed
+    channel: str = Field(default="all")
+    sheet: str | None = Field(default=None)
+    task_id: str | None = Field(default=None)
+    created_by: uuid.UUID | None = Field(default=None, foreign_key="users.id")
+    job_count: int = Field(default=0)
+    candidate_count: int = Field(default=0)
+    skipped_candidates: int = Field(default=0)
+    skipped_pairs: int = Field(default=0)
+    pair_count: int = Field(default=0)
+    done_count: int = Field(default=0)
+    entries: list | None = Field(default=None, sa_column=Column(JSONB))
+    summaries: list = Field(default_factory=list, sa_type=JSONB, sa_column_kwargs={"server_default": "[]"})
+    error: str | None = Field(default=None)
+    created_at: datetime = Field(default_factory=_now, sa_type=DateTime(timezone=True))
+    started_at: datetime | None = Field(default=None, sa_type=DateTime(timezone=True))
+    finished_at: datetime | None = Field(default=None, sa_type=DateTime(timezone=True))
+
+    @staticmethod
+    def _as_uuid(record_id) -> uuid.UUID | None:
+        if record_id in (None, ""):
+            return None
+        try:
+            return uuid.UUID(str(record_id))
+        except ValueError:
+            return None
+
+    @classmethod
+    async def get_by_id(cls, session: AsyncSession, record_id):
+        uid = cls._as_uuid(record_id)
+        if uid is None:
+            return None
+        result = await session.execute(select(cls).where(cls.id == uid))
+        return result.scalars().first()
+
+    @classmethod
+    async def get_active(cls, session: AsyncSession):
+        result = await session.execute(
+            select(cls)
+            .where(cls.status.in_(("queued", "running", "scoring")))
+            .order_by(cls.created_at.desc())
+        )
+        return result.scalars().first()
+
+    @classmethod
+    async def get_latest(cls, session: AsyncSession):
+        result = await session.execute(
+            select(cls).order_by(cls.created_at.desc()).limit(1)
+        )
+        return result.scalars().first()
+
+    @classmethod
+    async def insert_run(cls, session: AsyncSession, fields: dict, *, commit: bool = True):
+        row = cls(**fields)
+        session.add(row)
+        if commit:
+            await session.commit()
+            await session.refresh(row)
+        return row
+
+    @classmethod
+    async def update_run(cls, session: AsyncSession, record_id, fields: dict, *, commit: bool = True):
+        row = await cls.get_by_id(session, record_id)
+        if not row:
+            return None
+        for key, value in fields.items():
+            setattr(row, key, value)
+        session.add(row)
+        if commit:
+            await session.commit()
+            await session.refresh(row)
+        return row
+
+    @classmethod
+    async def append_summary(cls, session: AsyncSession, record_id, entry):
+        row = await cls.get_by_id(session, record_id)
+        if not row:
+            return None
+        items = list(row.summaries or [])
+        items.append(entry)
+        row.summaries = items
+        session.add(row)
+        await session.commit()
+        await session.refresh(row)
+        return row

backend/inbox/serializers.py

@@ -84,6 +84,7 @@ def serialize_message(message: Inbox_Messages, *, linkedin_url=None) -> dict:
         "resume_text": message.resume_text,
         "ats_score": message.ats_score,
         "ats_band": message.ats_band or None,
+        "professional_summary": message.professional_summary or None,
     }
 
 
@@ -93,6 +94,7 @@ def serialize_ats_result(row) -> dict:
         "job_post_id": str(row.job_post_id) if row.job_post_id else None,
         "overall_score": row.overall_score,
         "band": row.band or None,
+        "professional_summary": row.professional_summary or None,
         "computed_at": row.computed_at.isoformat() if row.computed_at else None,
     }
 
@@ -151,6 +153,7 @@ def serialize_application(message: Inbox_Messages, *, linkedin_url=None, light:
         "matched_at": message.matched_at.isoformat() if message.matched_at else None,
         "ats_score": message.ats_score,
         "ats_band": message.ats_band or None,
+        "professional_summary": message.professional_summary or None,
         "phone": message.candidate_phone_number,
         "experience": message.experience or "",
         "current_employment": message.current_employment or "",
@@ -212,3 +215,28 @@ def serialize_mailbox_sync_run(row) -> dict:
         "started_at": row.started_at.isoformat() if row.started_at else None,
         "finished_at": row.finished_at.isoformat() if row.finished_at else None,
     }
+
+
+def serialize_inbox_rescan_run(row) -> dict:
+    """Progress for the On-Hold catalogue ATS rescan, plus per-pair summaries."""
+    pair_count = int(row.pair_count or 0)
+    done_count = int(row.done_count or 0)
+    return {
+        "id": str(row.id),
+        "status": row.status,
+        "channel": row.channel,
+        "sheet": row.sheet,
+        "task_id": row.task_id,
+        "created_by": str(row.created_by) if row.created_by else None,
+        "job_count": int(row.job_count or 0),
+        "candidate_count": int(row.candidate_count or 0),
+        "skipped_candidates": int(row.skipped_candidates or 0),
+        "skipped_pairs": int(row.skipped_pairs or 0),
+        "pair_count": pair_count,
+        "done_count": done_count,
+        "summaries": list(row.summaries or []),
+        "error": row.error,
+        "created_at": row.created_at.isoformat() if row.created_at else None,
+        "started_at": row.started_at.isoformat() if row.started_at else None,
+        "finished_at": row.finished_at.isoformat() if row.finished_at else None,
+    }

backend/inbox/tasks.py

@@ -22,7 +22,7 @@ logger=logging.getLogger("inbox.tasks")
 _DONE=frozenset({"matched","skipped","no_text","failed","dlq"})
 
 
-async def score_message_against_job(record_id:str,job_id:str) -> dict:
+async def score_message_against_job(record_id:str,job_id:str,rescan_run_id=None) -> dict:
     """ATS-score one inbox CV against one job post — the no-upload path.
 
     The decoded attachment already on disk is the CV; the job post in the
@@ -57,7 +57,7 @@ async def score_message_against_job(record_id:str,job_id:str) -> dict:
             try:
                 # Attribute the rows to the job's owner — there is no request user
                 # in a background task.
-                results=await service.score_inbox(job_id,[record_id],{"id":str(job.created_by)})
+                results=await service.score_inbox(job_id,[record_id],{"id":str(job.created_by)},rescan_run_id=rescan_run_id)
             except HTTPException as exc:
                 # 400/404 from score_inbox are permanent (no attachment, bad ids);
                 # retrying cannot fix them.
@@ -138,6 +138,39 @@ async def score_inbox_message(record_id:str,job_id:str) -> dict:
     return await score_message_against_job(record_id,job_id)
 
 
+@broker.task(
+    task_name="inbox.rescan_on_hold",
+    retry_on_error=True,
+    max_retries=MAX_RETRIES,
+    delay=RETRY_DELAY,
+)
+async def rescan_on_hold_run(run_id:str,cursor:int=0) -> dict:
+    """Score On-Hold CVs against every job post, in short idempotent chunks."""
+    from g_sheet.scoring import score_form_against_job
+    from inbox.views import Email
+
+    async with session_scope() as session:
+        prepared=await Email(session=session).prepare_on_hold_rescan_chunk(run_id,cursor)
+    status=prepared.get("status")
+    if status in ("completed","failed","missing"):
+        return prepared
+    for item in prepared.get("batch") or []:
+        kind=item.get("kind")
+        record_id=item.get("record_id")
+        job_id=item.get("job_id")
+        try:
+            if kind=="form":
+                await score_form_against_job(record_id,job_id,run_id)
+            else:
+                await score_message_against_job(record_id,job_id,run_id)
+        except Exception as exc:
+            logger.warning("on-hold rescan failed for %s vs %s: %s",record_id,job_id,exc)
+    async with session_scope() as session:
+        return await Email(session=session).finish_on_hold_rescan_chunk(
+            run_id,prepared.get("next_cursor") or 0,bool(prepared.get("more")),
+        )
+
+
 @broker.task(
     task_name="inbox.match_message",
     retry_on_error=True,

backend/inbox/views.py

@@ -4,9 +4,9 @@ import uuid
 import httpx,os
 from fastapi import HTTPException
 from inbox.enums import Candidate_application_Status
-from inbox.models import Inbox_Messages,Inbox_Message_Triage,MailboxSyncRun,Inbox,SourceChannels,AtsResults
+from inbox.models import Inbox_Messages,Inbox_Message_Triage,MailboxSyncRun,InboxRescanRun,Inbox,SourceChannels,AtsResults
 from inbox.file_decoder import extract_pdf_attachments
-from inbox.serializers import serialize_application, serialize_message, serialize_triage, serialize_mailbox_sync_run, serialize_ats_result
+from inbox.serializers import serialize_application, serialize_message, serialize_triage, serialize_mailbox_sync_run, serialize_inbox_rescan_run, serialize_ats_result
 from inbox.plugins import (
     EMAIL_API_TOKEN,
     attach_email_pdfs_to_s3,
@@ -426,6 +426,240 @@ class Email:
             raise HTTPException(status_code=404,detail="No sync runs yet")
         return serialize_mailbox_sync_run(row)
 
+    async def start_on_hold_rescan(self,channel="all",sheet=None,current_user=None):
+        """Queue an On-Hold catalogue ATS scan against every job_posts row.
+
+        Returns an existing queued/running/scoring run instead of stacking another.
+        """
+        channel=(channel or "all").strip().lower()
+        if channel not in ("all","email","forms"):
+            raise HTTPException(status_code=422,detail="channel must be all, email, or forms")
+        sheet=(sheet or "").strip() or None
+        if channel!="forms":
+            sheet=None
+
+        active=await InboxRescanRun.get_active(self.session)
+        if active:
+            return serialize_inbox_rescan_run(active)
+
+        created_by=None
+        if isinstance(current_user,dict) and current_user.get("id"):
+            created_by=InboxRescanRun._as_uuid(current_user.get("id"))
+        row=await InboxRescanRun.insert_run(self.session,{
+            "status":"queued",
+            "channel":channel,
+            "sheet":sheet,
+            "created_by":created_by,
+        })
+        from inbox.tasks import rescan_on_hold_run
+        try:
+            task=await rescan_on_hold_run.kicker().with_labels(
+                created_at=datetime.now(timezone.utc).isoformat(),
+                correlation_id=str(row.id),
+                queue="inbox",
+            ).kiq(str(row.id),0)
+        except Exception as exc:
+            await InboxRescanRun.update_run(self.session,row.id,{
+                "status":"failed",
+                "error":str(exc),
+                "finished_at":datetime.now(timezone.utc),
+            })
+            raise HTTPException(status_code=503,detail="Could not queue On-Hold rescan") from exc
+        row=await InboxRescanRun.update_run(self.session,row.id,{"task_id":task.task_id})
+        return serialize_inbox_rescan_run(row)
+
+    async def get_on_hold_rescan(self,run_id=None):
+        if run_id:
+            row=await InboxRescanRun.get_by_id(self.session,run_id)
+            if not row:
+                raise HTTPException(status_code=404,detail="Rescan run not found")
+            return serialize_inbox_rescan_run(row)
+        row=await InboxRescanRun.get_active(self.session)
+        if row:
+            return serialize_inbox_rescan_run(row)
+        row=await InboxRescanRun.get_latest(self.session)
+        if not row:
+            return None
+        return serialize_inbox_rescan_run(row)
+
+    async def plan_on_hold_pairs(self,channel,sheet=None):
+        """Build (candidate, job) pairs that have never been ATS-scored.
+
+        Skip a candidate who already has a score against an active job.
+        Skip a pair that already exists on ats_results for that person/row.
+        """
+        from g_sheet.models import FormData
+        from job.job_post.models import JobPosts
+
+        job_ids=[str(jid) for jid in await JobPosts.list_ids(self.session)]
+        active_ids={str(jid) for jid in await JobPosts.list_ids(self.session,active_only=True)}
+        inbox_rows=[]
+        form_rows=[]
+        if channel in ("all","email"):
+            inbox_rows=await Inbox_Messages.list_on_hold_scan_rows(self.session)
+        if channel in ("all","forms"):
+            form_rows=await FormData.list_on_hold_scan_rows(self.session,sheet=sheet)
+
+        emails=[]
+        for row in inbox_rows:
+            if row.get("email"):
+                emails.append(row["email"])
+        for row in form_rows:
+            if row.get("email"):
+                emails.append(row["email"])
+        scored_by_email=await AtsResults.job_ids_by_emails(self.session,emails)
+        scored_by_message=await AtsResults.job_ids_for_messages(
+            self.session,[row["id"] for row in inbox_rows],
+        )
+        scored_by_form=await AtsResults.job_ids_for_forms(
+            self.session,[row["id"] for row in form_rows],
+        )
+        skipped_active=set()
+        for email,jobs in scored_by_email.items():
+            if jobs & active_ids:
+                skipped_active.add(email)
+        known_by_email={key:set(jobs) for key,jobs in scored_by_email.items()}
+
+        pairs=[]
+        skipped_candidates=0
+        skipped_pairs=0
+
+        def _already(email,row_jobs):
+            jobs=set(row_jobs or ())
+            if email:
+                jobs |= known_by_email.get(email) or set()
+            return jobs
+
+        def _mark(email,job_id):
+            if email:
+                known_by_email.setdefault(email,set()).add(job_id)
+
+        def _skip_for_active(email,row_jobs):
+            if email and email in skipped_active:
+                return True
+            if (row_jobs or set()) & active_ids:
+                return True
+            return False
+
+        for row in inbox_rows:
+            email=row.get("email")
+            mid=str(row["id"])
+            row_jobs=scored_by_message.get(mid) or set()
+            if not row.get("file_path"):
+                skipped_candidates += 1
+                continue
+            if _skip_for_active(email,row_jobs):
+                skipped_candidates += 1
+                continue
+            known=_already(email,row_jobs)
+            for job_id in job_ids:
+                if job_id in known:
+                    skipped_pairs += 1
+                    continue
+                pairs.append({"kind":"inbox","record_id":mid,"job_id":job_id})
+                known.add(job_id)
+                _mark(email,job_id)
+
+        for row in form_rows:
+            email=row.get("email")
+            fid=str(row["id"])
+            row_jobs=scored_by_form.get(fid) or set()
+            if _skip_for_active(email,row_jobs):
+                skipped_candidates += 1
+                continue
+            known=_already(email,row_jobs)
+            for job_id in job_ids:
+                if job_id in known:
+                    skipped_pairs += 1
+                    continue
+                pairs.append({"kind":"form","record_id":fid,"job_id":job_id})
+                known.add(job_id)
+                _mark(email,job_id)
+
+        return {
+            "job_count":len(job_ids),
+            "candidate_count":len(inbox_rows)+len(form_rows),
+            "skipped_candidates":skipped_candidates,
+            "skipped_pairs":skipped_pairs,
+            "pairs":pairs,
+        }
+
+    async def prepare_on_hold_rescan_chunk(self,run_id,cursor=0):
+        """Plan pairs if needed and return the next batch. Scoring happens outside."""
+        row=await InboxRescanRun.get_by_id(self.session,run_id)
+        if row is None:
+            return {"status":"missing"}
+        if row.status in ("failed","completed"):
+            return {"status":row.status}
+
+        now=datetime.now(timezone.utc)
+        entries=list(row.entries or [])
+        if not entries and int(cursor or 0)==0:
+            await InboxRescanRun.update_run(self.session,run_id,{
+                "status":"running",
+                "started_at":row.started_at or now,
+            })
+            plan=await self.plan_on_hold_pairs(row.channel,row.sheet)
+            entries=plan["pairs"]
+            await InboxRescanRun.update_run(self.session,run_id,{
+                "status":"scoring" if entries else "completed",
+                "job_count":plan["job_count"],
+                "candidate_count":plan["candidate_count"],
+                "skipped_candidates":plan["skipped_candidates"],
+                "skipped_pairs":plan["skipped_pairs"],
+                "pair_count":len(entries),
+                "done_count":0,
+                "entries":entries,
+                "finished_at":None if entries else now,
+            })
+            if not entries:
+                return {"status":"completed","pair_count":0,"batch":[]}
+
+        if not entries:
+            await InboxRescanRun.update_run(self.session,run_id,{
+                "status":"failed",
+                "error":"Rescan has no stored pairs",
+                "finished_at":datetime.now(timezone.utc),
+            })
+            return {"status":"failed","batch":[]}
+
+        chunk=4
+        start=int(cursor or 0)
+        batch=entries[start:start+chunk]
+        return {
+            "status":"scoring",
+            "batch":batch,
+            "pair_count":len(entries),
+            "next_cursor":min(start+len(batch),len(entries)),
+            "more":(start+len(batch))<len(entries),
+        }
+
+    async def finish_on_hold_rescan_chunk(self,run_id,next_cursor,more):
+        """Stamp progress and enqueue the next chunk after scores land."""
+        from inbox.tasks import rescan_on_hold_run
+
+        fields={
+            "status":"scoring" if more else "completed",
+            "done_count":int(next_cursor or 0),
+            "finished_at":None if more else datetime.now(timezone.utc),
+        }
+        await InboxRescanRun.update_run(self.session,run_id,fields)
+        if more:
+            try:
+                await rescan_on_hold_run.kicker().with_labels(
+                    created_at=datetime.now(timezone.utc).isoformat(),
+                    correlation_id=str(run_id),
+                    queue="inbox",
+                ).kiq(str(run_id),int(next_cursor or 0))
+            except Exception as exc:
+                await InboxRescanRun.update_run(self.session,run_id,{
+                    "status":"failed",
+                    "error":str(exc),
+                    "finished_at":datetime.now(timezone.utc),
+                })
+                raise
+        return {"status":fields["status"],"done_count":fields["done_count"]}
+
     async def run_mailbox_sync_page(self,top=100,skip=0,test_on=True,on_progress=None):
         """Pull one Outlook page, triage, ingest, enqueue matching. Returns summary.
 

backend/job/candidate/models.py

@@ -60,6 +60,7 @@ class Manual_UPLOAD_CANDIDATE(SQLModel, table=True):
     skills: list[str] = Field(default_factory=list, sa_type=JSONB, sa_column_kwargs={"server_default": "[]"})
     years_experience: int | None = Field(default=None)
     city: str | None = Field(default=None)
+    professional_summary: str | None = Field(default=None)
     reapplied: list[str] = Field(default_factory=list, sa_type=JSONB, sa_column_kwargs={"server_default": "[]"})
     education: str = Field(default="", sa_column_kwargs={"server_default": ""})
     # Why the CV is held (speculative / referral) and when retention expires.
@@ -377,6 +378,18 @@ class Manual_UPLOAD_CANDIDATE(SQLModel, table=True):
         result = await session.execute(select(cls).where(cls.id == uid))
         return result.scalars().first()
 
+    @classmethod
+    async def set_professional_summary(cls, session: AsyncSession, record_id, summary):
+        row = await cls.get_by_id(session, record_id)
+        if not row:
+            return None
+        row.professional_summary = (summary or "").strip() or None
+        row.updated_at = _now()
+        session.add(row)
+        await session.commit()
+        await session.refresh(row)
+        return row
+
     @classmethod
     async def get_by_ids(cls, session: AsyncSession, ids):
         keys = []
@@ -972,6 +985,7 @@ class Candidates(SQLModel, table=True):
     matched_keywords: list[str] = Field(default_factory=list, sa_type=JSON)
     missing_keywords: list[str] = Field(default_factory=list, sa_type=JSON)
     summary_critique: str | None = Field(default=None)
+    professional_summary: str | None = Field(default=None)
     # Public LinkedIn URL extracted from the scored CV. Fetch reads this column.
     linkedin_url: str | None = Field(default=None)
 

backend/job/candidate/plugins.py

@@ -139,6 +139,7 @@ def candidate_completed_fields(source, result):
         "matched_keywords": result.matched_keywords,
         "missing_keywords": result.missing_keywords,
         "summary_critique": result.summary_critique,
+        "professional_summary": result.professional_summary,
         "linkedin_url": None,
         "error_code": None,
         "error_message": None,

backend/job/candidate/serializers.py

@@ -30,6 +30,7 @@ def serialize_candidate(row) -> dict:
         "matched_keywords": list(row.matched_keywords or []),
         "missing_keywords": list(row.missing_keywords or []),
         "summary_critique": row.summary_critique,
+        "professional_summary": row.professional_summary or None,
         "linkedin_url": row.linkedin_url or None,
         "status": row.status,
         "error_code": row.error_code,
@@ -208,6 +209,10 @@ def serialize_candidate_profile(
         "match_status": message.match_status if message else None,
         "match_error": message.match_error if message else None,
         "matched_at": message.matched_at.isoformat() if message and message.matched_at else None,
+        "professional_summary": (
+            (message.professional_summary if message else None)
+            or (user.professional_summary if user else None)
+        ),
         "file_path": _first_file_path(message.file_path if message else None),
         "job_posts": [],
     }
@@ -278,6 +283,7 @@ def serialize_manual_candidate_profile(row, user, job_post) -> Dict[str, Any]:
         "match_status": None,
         "match_error": None,
         "matched_at": None,
+        "professional_summary": row.professional_summary or (user.professional_summary if user else None),
         "job_posts": [job_payload] if job_payload else [],
         "favorite": None,
         "rating": None,

backend/job/candidate/views.py

@@ -9,7 +9,7 @@ from app.core.errors import ATSError,ErrorCode
 from app.models.scoring import CompletedCandidate
 from app.services.pdf import extract_resume,sanitize_filename
 from app.services.scoring import score_batch
-from inbox.models import Inbox_Messages,Inbox,Inbox_Message_Triage,AtsResults
+from inbox.models import Inbox_Messages,Inbox,Inbox_Message_Triage,InboxRescanRun,AtsResults
 from job.candidate.models import Candidates
 from job.candidate.plugins import (
     FILE_NOT_FOUND,
@@ -632,7 +632,7 @@ class CandidateScoring:
             sources.append(source)
         return await self._score_and_persist(job_id,sources,"upload",current_user)
 
-    async def score_inbox(self,job_id,message_ids,current_user):
+    async def score_inbox(self,job_id,message_ids,current_user,rescan_run_id=None):
         """Score PDF attachments of inbox messages (S3 URLs or legacy local paths)."""
         from inbox.plugins import load_file_bytes
 
@@ -671,7 +671,7 @@ class CandidateScoring:
                 sources.append(source)
         if not sources:
             raise HTTPException(status_code=400,detail="No attachments found for the given message(s)")
-        return await self._score_and_persist(job_id,sources,"inbox",current_user)
+        return await self._score_and_persist(job_id,sources,"inbox",current_user,rescan_run_id=rescan_run_id)
 
     async def score_bank(self,job_id,record_ids,current_user):
         """ATS-score CVs already sitting in the bank — tier 2, the paid step.
@@ -747,7 +747,7 @@ class CandidateScoring:
         data=serialize_candidate(row)
         return await CandidateView(session=self.session).attach_application_history(data)
 
-    async def _score_and_persist(self,job_id,sources,source_kind,current_user):
+    async def _score_and_persist(self,job_id,sources,source_kind,current_user,rescan_run_id=None):
         job=await JobPosts.get_job_post_by_id(self.session,job_id)
         if job is None or job.is_deleted:
             raise HTTPException(status_code=404,detail="Job post not found")
@@ -790,7 +790,7 @@ class CandidateScoring:
                 ):
                     await self.session.commit()
             rows.append(row)
-        await self._sync_ats_results(source_kind,job,rows,sources,current_user)
+        await self._sync_ats_results(source_kind,job,rows,sources,current_user,rescan_run_id=rescan_run_id)
         rows.sort(key=lambda r:(0,-(r.match_score or 0)) if r.status=="completed" else (1,0))
         return [serialize_candidate(row) for row in rows]
 
@@ -833,7 +833,8 @@ class CandidateScoring:
                 fields_by_slot[slot]=candidate_failed_fields(source,result.error_code,result.error_message)
         return fields_by_slot
 
-    async def _sync_ats_results(self,source_kind,job,rows,sources,current_user=None):
+    async def _sync_ats_results(self,source_kind,job,rows,sources,current_user=None,rescan_run_id=None):
+        run_id=self._as_rescan_run_id(rescan_run_id)
         if source_kind=="inbox":
             best={}
             for source,row in zip(sources,rows):
@@ -844,21 +845,50 @@ class CandidateScoring:
                         best[mid]=row
             for message_id,row in best.items():
                 try:
-                    await self._sync_inbox_ats(message_id,job,row,current_user=current_user)
+                    await self._sync_inbox_ats(message_id,job,row,current_user=current_user,rescan_run_id=run_id)
                 except Exception:
                     await self.session.rollback()
                     logger.exception("inbox ATS denorm failed for message %s",message_id)
             return
-        for row in rows:
+        for source,row in zip(sources,rows):
             if row.status!="completed":
                 continue
             try:
-                await self._sync_upload_ats(job,row,current_user=current_user)
+                await self._sync_upload_ats(job,row,source,current_user=current_user,rescan_run_id=run_id)
             except Exception:
                 await self.session.rollback()
                 logger.exception("upload ATS history failed for candidate %s",row.id)
 
-    async def _sync_inbox_ats(self,message_id,job,row,current_user=None):
+    @staticmethod
+    def _as_rescan_run_id(value):
+        if value in (None,""):
+            return None
+        try:
+            return uuid.UUID(str(value))
+        except (TypeError,ValueError):
+            return None
+
+    @staticmethod
+    def _summary_text(row):
+        return (getattr(row,"professional_summary",None) or "").strip() or None
+
+    async def _stamp_identity_summary(self,row,summary):
+        identity=await AtsResults.resolve_identity(self.session,row.candidate_email,row.id)
+        if identity.get("user_id"):
+            await Users.set_professional_summary(self.session,identity["user_id"],summary)
+        return identity
+
+    async def _append_rescan_summary(self,rescan_run_id,kind,record_id,job,summary):
+        if not rescan_run_id:
+            return
+        await InboxRescanRun.append_summary(self.session,rescan_run_id,{
+            "kind":kind,
+            "record_id":str(record_id),
+            "job_post_id":str(job.id),
+            "professional_summary":summary,
+        })
+
+    async def _sync_inbox_ats(self,message_id,job,row,current_user=None,rescan_run_id=None):
         """Land a completed score on inbox_messages / inbox / ats_results.
 
         message_id is the scoring call's known inbox_messages PK, not a column
@@ -868,6 +898,7 @@ class CandidateScoring:
         if msg is None:
             return
         band=CandidateView._recommendation(row.match_score) or ""
+        summary=self._summary_text(row)
         denorm=True
         assigned=msg.assigned_job_post_id
         link=await Inbox.get_inbox_by_message_id(self.session,message_id)
@@ -876,11 +907,12 @@ class CandidateScoring:
             denorm=existing is None
         if denorm:
             await Inbox_Messages.set_ats_score(self.session,message_id,row.match_score,band)
+        await Inbox_Messages.set_professional_summary(self.session,message_id,summary)
+        identity=await self._stamp_identity_summary(row,summary)
         if link is None:
             return
         old=await AtsResults.get_current_for_inbox(self.session,link.id)
         old_score=old.overall_score if old else None
-        identity=await AtsResults.resolve_identity(self.session,row.candidate_email,row.id)
         await AtsResults.insert_result(self.session,{
             "inbox_id":link.id,
             **identity,
@@ -888,8 +920,11 @@ class CandidateScoring:
             "overall_score":float(row.match_score),
             "band":band,
             "model_name":row.model,
+            "professional_summary":summary,
+            "rescan_run_id":rescan_run_id,
             "is_current":True,
         })
+        await self._append_rescan_summary(rescan_run_id,"email",message_id,job,summary)
         await HistoryRecorder(self.session).record(
             HistoryEvent.ATS_SCORED.value,
             current_user=current_user,user_id=link.user_id,inbox_id=link.id,
@@ -898,10 +933,18 @@ class CandidateScoring:
             description=f"{band} against {job.title or 'job'}",commit=True,
         )
 
-    async def _sync_upload_ats(self,job,row,current_user=None):
+    async def _sync_upload_ats(self,job,row,source=None,current_user=None,rescan_run_id=None):
         """History row for an upload-sourced score — inbox_id stays NULL."""
         band=CandidateView._recommendation(row.match_score) or ""
-        identity=await AtsResults.resolve_identity(self.session,row.candidate_email,row.id)
+        summary=self._summary_text(row)
+        identity=await self._stamp_identity_summary(row,summary)
+        mid=(source or {}).get("manual_upload_candidate_id")
+        if not mid and row.candidate_email:
+            manual=await Manual_UPLOAD_CANDIDATE.get_by_email_and_job(self.session,row.candidate_email,job.id)
+            if manual:
+                mid=manual.id
+        if mid:
+            await Manual_UPLOAD_CANDIDATE.set_professional_summary(self.session,mid,summary)
         old=None
         if identity.get("user_id"):
             old=await AtsResults.get_current_for_user(self.session,identity["user_id"],job.id)
@@ -915,8 +958,11 @@ class CandidateScoring:
             "overall_score":float(row.match_score),
             "band":band,
             "model_name":row.model,
+            "professional_summary":summary,
+            "rescan_run_id":rescan_run_id,
             "is_current":True,
         })
+        await self._append_rescan_summary(rescan_run_id,"upload",mid or row.id,job,summary)
         await HistoryRecorder(self.session).record(
             HistoryEvent.ATS_SCORED.value,
             current_user=current_user,user_id=identity.get("user_id"),
@@ -1453,6 +1499,8 @@ class CandidateView:
             base["candidate_id"]=str(chosen.candidate_id) if chosen.candidate_id else None
             if chosen.user_id and not base.get("user_id"):
                 base["user_id"]=str(chosen.user_id)
+            if chosen.professional_summary and not base.get("professional_summary"):
+                base["professional_summary"]=chosen.professional_summary
             scored=None
             if chosen.candidate_id:
                 scored=await Candidates.get_candidate_by_id(self.session,str(chosen.candidate_id))
@@ -1464,6 +1512,8 @@ class CandidateView:
                 base["matched_keywords"]=list(scored.matched_keywords or [])
                 base["missing_keywords"]=list(scored.missing_keywords or [])
                 base["summary_critique"]=scored.summary_critique
+                if scored.professional_summary and not base.get("professional_summary"):
+                    base["professional_summary"]=scored.professional_summary
         else:
             for record in records:
                 score,band=self._score_from_message(record)

backend/job/job_post/models.py

@@ -109,6 +109,15 @@ class JobPosts(SQLModel, table=True):
         )
         return result.scalars().all()
 
+    @classmethod
+    async def list_ids(cls, session: AsyncSession, *, active_only: bool = False):
+        """Non-deleted job_posts.id values. active_only limits to live openings."""
+        statement = select(cls.id).where(cls.is_deleted == False)  # noqa: E712
+        if active_only:
+            statement = statement.where(cls.is_active == True)  # noqa: E712
+        result = await session.execute(statement)
+        return [row[0] for row in result.all()]
+
     @classmethod
     async def get_by_ids(cls, session: AsyncSession, ids: list[str], *, active_only: bool = True):
         uids = []

backend/migrations/manual/032_inbox_rescan_runs.sql

@@ -0,0 +1,26 @@
+-- 032_inbox_rescan_runs.sql
+-- On-Hold catalogue ATS rescan: one run row so the Inbox ReScan button can
+-- poll progress while workers score (candidate, job_post) pairs.
+
+CREATE TABLE IF NOT EXISTS app.inbox_rescan_runs (
+    id uuid PRIMARY KEY,
+    status varchar NOT NULL DEFAULT 'queued',
+    channel varchar NOT NULL DEFAULT 'all',
+    sheet varchar,
+    task_id varchar,
+    created_by uuid REFERENCES app.users (id),
+    job_count integer NOT NULL DEFAULT 0,
+    candidate_count integer NOT NULL DEFAULT 0,
+    skipped_candidates integer NOT NULL DEFAULT 0,
+    skipped_pairs integer NOT NULL DEFAULT 0,
+    pair_count integer NOT NULL DEFAULT 0,
+    done_count integer NOT NULL DEFAULT 0,
+    entries jsonb,
+    error text,
+    created_at timestamptz NOT NULL DEFAULT now(),
+    started_at timestamptz,
+    finished_at timestamptz
+);
+
+CREATE INDEX IF NOT EXISTS ix_inbox_rescan_runs_status
+    ON app.inbox_rescan_runs (status);

backend/migrations/manual/033_professional_summary.sql

@@ -0,0 +1,30 @@
+-- 033_professional_summary.sql
+-- Job-agnostic professional summary (tech-stack speciality + department).
+-- Latest denorm on each intake row; per-run snapshot on ats_results.
+-- ats_results.rescan_run_id is the 1:N link from an On-Hold ReScan run.
+-- Applied at startup by alembic_setup.run_manual_sql().
+
+ALTER TABLE app.inbox_messages
+    ADD COLUMN IF NOT EXISTS professional_summary TEXT;
+
+ALTER TABLE app.form_data
+    ADD COLUMN IF NOT EXISTS professional_summary TEXT;
+
+ALTER TABLE app.manual_upload_candidate
+    ADD COLUMN IF NOT EXISTS professional_summary TEXT;
+
+ALTER TABLE app.users
+    ADD COLUMN IF NOT EXISTS professional_summary TEXT;
+
+ALTER TABLE app.candidates
+    ADD COLUMN IF NOT EXISTS professional_summary TEXT;
+
+ALTER TABLE app.ats_results
+    ADD COLUMN IF NOT EXISTS professional_summary TEXT,
+    ADD COLUMN IF NOT EXISTS rescan_run_id UUID REFERENCES app.inbox_rescan_runs (id);
+
+ALTER TABLE app.inbox_rescan_runs
+    ADD COLUMN IF NOT EXISTS summaries JSONB NOT NULL DEFAULT '[]'::jsonb;
+
+CREATE INDEX IF NOT EXISTS ix_ats_results_rescan_run_id
+    ON app.ats_results (rescan_run_id);

backend/tests/test_g_sheet_scoring.py

@@ -34,12 +34,14 @@ def test_serialize_form_ats_shape():
         job_post_id = job_id
         overall_score = 81.0
         band = "Potential Match"
+        professional_summary = "Python backend in Engineering."
         computed_at = None
 
     assert serialize_form_ats(Row()) == {
         "job_post_id": str(job_id),
         "overall_score": 81.0,
         "band": "Potential Match",
+        "professional_summary": "Python backend in Engineering.",
         "computed_at": None,
     }
 

backend/tests/test_professional_summary.py

@@ -0,0 +1,83 @@
+"""Professional summary — persist fields and serializer shape. No DB."""
+
+from types import SimpleNamespace
+from uuid import uuid4
+
+from g_sheet.scoring import serialize_form_ats
+from inbox.serializers import serialize_ats_result, serialize_inbox_rescan_run
+from job.candidate.plugins import candidate_completed_fields, candidate_failed_fields
+
+
+def test_failed_fields_omit_professional_summary():
+    source = {"safe_name": "a.pdf", "file_path": None, "sha256": "x"}
+    fields = candidate_failed_fields(source, "PDF_TEXT_UNAVAILABLE", "empty")
+    assert "professional_summary" not in fields
+
+
+def test_completed_fields_include_professional_summary():
+    result = SimpleNamespace(
+        candidate_name="Ada",
+        job_title="Engineer",
+        current_company="Acme",
+        years_experience=6,
+        match_score=82,
+        matched_keywords=["Python"],
+        missing_keywords=["AWS"],
+        summary_critique="Strong Python, missing cloud.",
+        professional_summary="Python backend specialist in Engineering.",
+    )
+    source = {"safe_name": "a.pdf", "file_path": None, "sha256": "x"}
+    fields = candidate_completed_fields(source, result)
+    assert fields["professional_summary"] == "Python backend specialist in Engineering."
+
+
+def test_serialize_ats_result_includes_summary():
+    job_id = uuid4()
+    row = SimpleNamespace(
+        job_post_id=job_id,
+        overall_score=81.0,
+        band="Potential Match",
+        professional_summary="Python backend in Engineering.",
+        computed_at=None,
+    )
+    assert serialize_ats_result(row) == {
+        "job_post_id": str(job_id),
+        "overall_score": 81.0,
+        "band": "Potential Match",
+        "professional_summary": "Python backend in Engineering.",
+        "computed_at": None,
+    }
+    assert serialize_form_ats(row) == serialize_ats_result(row)
+
+
+def test_serialize_inbox_rescan_run_includes_summaries():
+    run_id = uuid4()
+    entry = {
+        "kind": "email",
+        "record_id": str(uuid4()),
+        "job_post_id": str(uuid4()),
+        "professional_summary": "Python backend in Engineering.",
+    }
+    row = SimpleNamespace(
+        id=run_id,
+        status="completed",
+        channel="all",
+        sheet=None,
+        task_id=None,
+        created_by=None,
+        job_count=2,
+        candidate_count=1,
+        skipped_candidates=0,
+        skipped_pairs=0,
+        pair_count=1,
+        done_count=1,
+        summaries=[entry],
+        error=None,
+        created_at=None,
+        started_at=None,
+        finished_at=None,
+    )
+    payload = serialize_inbox_rescan_run(row)
+    assert payload["summaries"] == [entry]
+    assert payload["id"] == str(run_id)
+    assert payload["done_count"] == 1

backend/users/models.py

@@ -62,6 +62,7 @@ class Users(SQLModel, table=True):
     # mentions LinkedIn; never overwrite a stored value with empty.
     linkedin_url: str | None = Field(default=None)
     city: str | None = Field(default=None)
+    professional_summary: str | None = Field(default=None)
     # Prior job_post_ids for this email across candidate tables. [] until a
     # later application finds an already-linked job.
     reapplied: list[str] = Field(default_factory=list, sa_type=JSONB, sa_column_kwargs={"server_default": "[]"})
@@ -354,5 +355,20 @@ class Users(SQLModel, table=True):
         await session.commit()
         return updated
 
+    @classmethod
+    async def set_professional_summary(cls, session: AsyncSession, user_id, summary):
+        uid = cls._as_uuid(user_id)
+        if uid is None:
+            return None
+        user = await cls.get_user_id(session, uid)
+        if user is None:
+            return None
+        user.professional_summary = (summary or "").strip() or None
+        user.updated_at = _now()
+        session.add(user)
+        await session.commit()
+        await session.refresh(user)
+        return user
+
 
 import job.candidate.models as _candidate_models  # noqa: E402, F401

frontend/src/api/inbox.js

@@ -81,6 +81,19 @@ export function getMailboxSync(runId) {
   return request('/email/sync/fetch', { params: { run_id: runId } })
 }
 
+/** Score every On-Hold CV against all job posts. Returns a run immediately. */
+export function startOnHoldRescan({ channel = 'all', sheet } = {}) {
+  return request('/inbox/rescan-on-hold', {
+    method: 'POST',
+    body: { channel, sheet },
+  })
+}
+
+/** Poll one On-Hold rescan (or the active/latest run when runId is omitted). */
+export function getOnHoldRescan(runId) {
+  return request('/inbox/rescan-on-hold', { params: { run_id: runId } })
+}
+
 /**
  * Legacy synchronous sync — blocks until the page is ingested. Prefer
  * startMailboxSync + getMailboxSync so closing the tab cannot kill the job.

frontend/src/lib/queryKeys.js

@@ -29,6 +29,7 @@ export const qk = {
     // override or a sync needs no extra invalidation.
     triage: (p = {}) => ['mailbox', 'triage', p],
     sync: (id) => ['mailbox', 'sync', id],
+    rescan: (id) => ['mailbox', 'rescan', id],
     // Sheet form applicants live under the same mailbox prefix so the Inbox
     // channel toggle can invalidate both email and form caches together.
     formSheets: () => ['mailbox', 'form-sheets'],

frontend/src/screens/CandidateProfile.jsx

@@ -316,14 +316,19 @@ export default function CandidateProfile({
         {/* No score anywhere -> the whole block goes, rather than a ring drawn
             around a blank. Seed-backed callers still pass a number and are
             unaffected; only live candidates the engine never scored drop out. */}
-        {(atsScore ?? live?.ai_score ?? c.aiScore) != null && (
-          <div style={{ textAlign: 'center' }}>
-            <ScoreChip score={atsScore ?? live?.ai_score ?? c.aiScore} />
-            {/* The band caption replaces the static label only when the caller
-                actually fetched one — every other screen keeps "AI Match". */}
-            <div className="cell-sub" style={{ marginTop: 4 }}>{recommendation || 'AI Match'}</div>
+        {(atsScore ?? live?.ai_score ?? c.aiScore) != null || live?.professional_summary ? (
+          <div className="flex items-center gap-12" style={{ maxWidth: 380 }}>
+            {(atsScore ?? live?.ai_score ?? c.aiScore) != null && (
+              <div style={{ textAlign: 'center', flex: '0 0 auto' }}>
+                <ScoreChip score={atsScore ?? live?.ai_score ?? c.aiScore} />
+                {/* The band caption replaces the static label only when the caller
+                    actually fetched one — every other screen keeps "AI Match". */}
+                <div className="cell-sub" style={{ marginTop: 4 }}>{recommendation || 'AI Match'}</div>
+              </div>
+            )}
+            {live?.professional_summary ? <div className="ats-summary">{live.professional_summary}</div> : null}
           </div>
-        )}
+        ) : null}
       </div>
 
       <div style={{ marginTop: 22 }}>
@@ -346,6 +351,7 @@ export default function CandidateProfile({
               <Info label="Current Company" val={live.currentCompany} />
               <Info label="Experience" val={live.experience} />
               <Info label="Education" val={live.education} />
+              <Info label="Professional Summary" val={live.professional_summary} />
               <Info label="Source" val={live.source} />
               <Info label="Recruiter" val={live.recruiter} />
               <Info label="Applied On" val={fmtWhen(live.applied)} />

frontend/src/screens/Inbox.jsx

@@ -324,6 +324,23 @@ function emailAtsScore(row) {
     ?? (fromJobs.length ? Math.max(...fromJobs) : null)
 }
 
+function AtsHero({ score, band, ringColor, summary }) {
+  if (score == null && !summary) return null
+  return (
+    <div className="flex items-center gap-12" style={{ maxWidth: 380 }}>
+      {score != null && (
+        <div style={{ textAlign: 'center', flex: '0 0 auto' }}>
+          <div className="ats-ring" style={{ width: 84, height: 84, '--pct': score, '--c': ringColor }}>
+            <div className="ats-val"><div className="ats-num" style={{ fontSize: 22 }}>{Math.round(score)}</div></div>
+          </div>
+          <div className="cell-sub" style={{ marginTop: 4 }}>{band || 'ATS Score'}</div>
+        </div>
+      )}
+      {summary ? <div className="ats-summary">{summary}</div> : null}
+    </div>
+  )
+}
+
 /** Assigned job title for export — list rows may carry the post or only an id + jobPosts. */
 function assignedJobTitle(row) {
   const fromPost = row?.assignedPost?.title
@@ -418,6 +435,7 @@ function mapFormRow(row) {
       : [],
     atsResults,
     atsScore,
+    professionalSummary: (row.professional_summary || '').trim() || '',
     assignedId,
     assignedPost: row.assigned_job_post || null,
     isReapplicant: Boolean(row.is_reapplicant),
@@ -565,6 +583,7 @@ async function fetchMessageDetail(recordId) {
     assignedPost: row.assigned_job_post || null,
     atsResults: Array.isArray(row.ats_results) ? row.ats_results : [],
     atsScore: emailAtsScore(row),
+    professionalSummary: (row.professional_summary || '').trim() || '',
     isReapplicant: Boolean(row.is_reapplicant),
     previousApplications: Array.isArray(row.previous_applications) ? row.previous_applications : [],
   }
@@ -623,6 +642,7 @@ async function fetchApplications(params) {
         assignedPost: row.assigned_job_post || null,
         atsResults: Array.isArray(row.ats_results) ? row.ats_results : [],
         isReapplicant: Boolean(row.is_reapplicant),
+        professionalSummary: (row.professional_summary || '').trim() || '',
         previousApplications: Array.isArray(row.previous_applications) ? row.previous_applications : [],
       }
     }),
@@ -1093,7 +1113,7 @@ function InboxFilters({ filters, cities, sources, showLinkFilters, active, open,
  * Owns its own interval so the minute count stays current without re-rendering
  * the queue: under "Show all" that parent render is a thousand rows.
  */
-function QueueFreshness({ at, refreshing, onRefresh }) {
+function QueueFreshness({ at, refreshing, onRefresh, rescan }) {
   const [, tick] = useState(0)
   useEffect(() => {
     if (!at) return undefined
@@ -1101,24 +1121,44 @@ function QueueFreshness({ at, refreshing, onRefresh }) {
     return () => clearInterval(t)
   }, [at])
 
+  const rescanning = Boolean(rescan?.busy)
+  const progress = rescan?.progress
+
   return (
     <div className="inbox-freshness">
       <span className="cell-sub">
-        {refreshing
-          ? 'Updating…'
-          : at
-            ? `Updated ${agoLabel(at)}`
-            : 'Loading…'}
+        {rescanning
+          ? (progress
+            ? `ReScan ${progress.done}/${progress.total}…`
+            : 'ReScan in progress…')
+          : refreshing
+            ? 'Updating…'
+            : at
+              ? `Updated ${agoLabel(at)}`
+              : 'Loading…'}
       </span>
-      <button
-        type="button"
-        className="btn btn-ghost btn-sm"
-        onClick={onRefresh}
-        disabled={refreshing}
-        title="Re-read the applications already saved here. Use Sync to pull new mail from Outlook."
-      >
-        <Icon name="refresh" /> Refresh
-      </button>
+      <div className="inbox-freshness-actions">
+        {rescan?.visible && (
+          <button
+            type="button"
+            className="btn btn-ghost btn-sm"
+            onClick={rescan.onClick}
+            disabled={rescanning || !rescan.enabled}
+            title="Score every On-Hold CV against all job posts. Skips candidates already scored for an active job, and any candidate+job pair already in the system."
+          >
+            <Icon name="sparkles" /> {rescanning ? 'Scanning…' : 'ReScan'}
+          </button>
+        )}
+        <button
+          type="button"
+          className="btn btn-ghost btn-sm"
+          onClick={onRefresh}
+          disabled={refreshing || rescanning}
+          title="Re-read the applications already saved here. Use Sync to pull new mail from Outlook."
+        >
+          <Icon name="refresh" /> Refresh
+        </button>
+      </div>
     </div>
   )
 }
@@ -1736,6 +1776,70 @@ export default function Inbox() {
     return undefined
   }, [qc, syncRun.data, toast])
 
+  const onHoldTab = tab === 'On-Hold'
+  const [rescanRunId, setRescanRunId] = useState(null)
+  const rescanToastShown = useRef(null)
+
+  const rescan = useMutation({
+    mutationFn: () => inboxApi.startOnHoldRescan({
+      channel,
+      sheet: channel === 'forms' ? formSheet : undefined,
+    }),
+    onSuccess: (res) => {
+      const id = res?.data?.id
+      if (id) setRescanRunId(id)
+      toast('ReScan started — On-Hold CVs will be scored against every job post', 'info')
+    },
+    onError: (err) => toast(friendlyAuthError(err, 'ReScan failed'), 'error'),
+  })
+
+  const rescanRun = useQuery({
+    queryKey: qk.mailbox.rescan(rescanRunId || 'latest'),
+    queryFn: async () => {
+      const res = await inboxApi.getOnHoldRescan(rescanRunId)
+      return res?.data ?? null
+    },
+    enabled: onHoldTab,
+    refetchInterval: (q) => {
+      const status = q.state.data?.status
+      return status === 'queued' || status === 'running' || status === 'scoring' ? 1500 : false
+    },
+  })
+
+  const rescanBusy = ['queued', 'running', 'scoring'].includes(rescanRun.data?.status)
+    || rescan.isPending
+
+  useEffect(() => {
+    const run = rescanRun.data
+    if (!run?.id) return undefined
+    const active = ['queued', 'running', 'scoring'].includes(run.status)
+    if (active) {
+      setRescanRunId((id) => id || run.id)
+      const timer = setInterval(() => {
+        qc.invalidateQueries({ queryKey: qk.mailbox.all() })
+      }, 8000)
+      return () => clearInterval(timer)
+    }
+    if (rescanRunId !== run.id) return undefined
+    if (run.status === 'completed' && rescanToastShown.current !== run.id) {
+      rescanToastShown.current = run.id
+      const scored = run.done_count ?? 0
+      const skipped = (run.skipped_candidates ?? 0) + (run.skipped_pairs ?? 0)
+      toast(
+        scored
+          ? `ReScan finished — ${scored} ATS score${scored === 1 ? '' : 's'} saved${skipped ? `, ${skipped} skipped` : ''}`
+          : `ReScan finished — nothing new to score${skipped ? ` (${skipped} skipped)` : ''}`,
+        'success',
+      )
+      qc.invalidateQueries({ queryKey: qk.mailbox.all() })
+    }
+    if (run.status === 'failed' && rescanToastShown.current !== run.id) {
+      rescanToastShown.current = run.id
+      toast(run.error || 'ReScan failed', 'error')
+    }
+    return undefined
+  }, [qc, rescanRun.data, rescanRunId, toast])
+
   return (
     <div className="page">
       <PageHeader
@@ -1856,7 +1960,20 @@ export default function Inbox() {
                   onChange={setInboxFilter}
                   onClear={() => { setInboxFilters(EMPTY_INBOX_FILTERS); setSkip(0); setSelectedId(null) }}
                 />
-                <QueueFreshness at={updatedAt} refreshing={refreshing} onRefresh={refreshQueue} />
+                <QueueFreshness
+                  at={updatedAt}
+                  refreshing={refreshing}
+                  onRefresh={refreshQueue}
+                  rescan={onHoldTab ? {
+                    visible: true,
+                    enabled: canEdit,
+                    busy: rescanBusy,
+                    onClick: () => rescan.mutate(),
+                    progress: rescanRun.data?.pair_count
+                      ? { done: rescanRun.data.done_count || 0, total: rescanRun.data.pair_count }
+                      : null,
+                  } : null}
+                />
               </div>
               <div className="inbox-list-body">
                 {/* Hairline, not a skeleton: the rows below stay readable and
@@ -2247,14 +2364,14 @@ function FormApplicantDetail({
             <div className="fw-600">{i.rowNumber}</div>
           </div>
         )}
-        {selectedAts != null && (
-          <div style={{ textAlign: 'center' }}>
-            <div className="ats-ring" style={{ width: 84, height: 84, '--pct': selectedAts.score, '--c': selectedRingColor }}>
-              <div className="ats-val"><div className="ats-num" style={{ fontSize: 22 }}>{Math.round(selectedAts.score)}</div></div>
-            </div>
-            <div className="cell-sub" style={{ marginTop: 4 }}>{selectedAts.band || 'ATS Score'}</div>
-          </div>
-        )}
+        {selectedAts != null || i.professionalSummary ? (
+          <AtsHero
+            score={selectedAts?.score}
+            band={selectedAts?.band}
+            ringColor={selectedRingColor}
+            summary={i.professionalSummary}
+          />
+        ) : null}
       </div>
 
       <PreviousApplications row={i} />
@@ -2622,14 +2739,14 @@ function ApplicationDetail({
             {loading && <span className="cell-sub">Loading details…</span>}
           </div>
         </div>
-        {i.atsScore != null && (
-          <div style={{ textAlign: 'center' }}>
-            <div className="ats-ring" style={{ width: 84, height: 84, '--pct': i.atsScore, '--c': ringColor }}>
-              <div className="ats-val"><div className="ats-num" style={{ fontSize: 22 }}>{Math.round(i.atsScore)}</div></div>
-            </div>
-            <div className="cell-sub" style={{ marginTop: 4 }}>ATS Score</div>
-          </div>
-        )}
+        {i.atsScore != null || i.professionalSummary ? (
+          <AtsHero
+            score={i.atsScore}
+            band="ATS Score"
+            ringColor={ringColor}
+            summary={i.professionalSummary}
+          />
+        ) : null}
       </div>
 
       <PreviousApplications row={i} />

frontend/src/styles/styles.css

@@ -1173,6 +1173,12 @@ canvas { width: 100%; max-width: 100%; display: block; }
   font-size: 12px;
   flex-shrink: 0;
 }
+.inbox-freshness-actions {
+  display: flex;
+  align-items: center;
+  gap: 6px;
+  flex-shrink: 0;
+}
 /* Sheet Forms link filters. Deliberately NOT .filter-panel: that grid is four
    columns wide and its breakpoints watch the viewport, but this lives in a
    minmax(280px, 34%) column, so on a wide screen it would pack four columns into
@@ -1319,6 +1325,7 @@ canvas { width: 100%; max-width: 100%; display: block; }
 .ats-ring .ats-val { position: relative; z-index: 1; text-align: center; }
 .ats-ring .ats-num { font-family: var(--font-display); font-size: 30px; font-weight: 700; letter-spacing: -0.02em; line-height: 1; }
 .ats-ring .ats-lbl { font-size: 11px; color: var(--text-3); font-weight: 600; }
+.ats-summary { font-size: var(--fs-xs); color: var(--text-2); line-height: 1.4; max-width: 280px; display: -webkit-box; -webkit-line-clamp: 3; -webkit-box-orient: vertical; overflow: hidden; }
 .skill-pill { display: inline-flex; align-items: center; gap: 5px; font-size: var(--fs-xs); font-weight: 600; padding: 4px 10px; border-radius: var(--radius-sm); }
 .skill-pill svg { width: 12px; height: 12px; }
 .skill-matched { background: var(--success-soft); color: var(--success); }

tests/integration/test_api.py

@@ -62,6 +62,7 @@ class TestHappyPath:
             "matched_keywords",
             "missing_keywords",
             "summary_critique",
+            "professional_summary",
         }
         assert result["status"] == "completed"
 

tests/unit/test_models.py

@@ -82,12 +82,14 @@ class TestProfileFields:
         assert result.job_title is None
         assert result.current_company is None
         assert result.years_experience is None
+        assert result.professional_summary is None
 
     def test_blank_strings_normalise_to_none(self) -> None:
-        result = score(candidate_name="   ", job_title="", current_company="\n\t")
+        result = score(candidate_name="   ", job_title="", current_company="\n\t", professional_summary="  ")
         assert result.candidate_name is None
         assert result.job_title is None
         assert result.current_company is None
+        assert result.professional_summary is None
 
     def test_whitespace_is_collapsed(self) -> None:
         result = score(candidate_name="  Ada   Lovelace ", job_title="Senior\nEngineer")
@@ -107,6 +109,14 @@ class TestProfileFields:
         with pytest.raises(ValidationError):
             score(candidate_name="x" * 121)
 
+    def test_professional_summary_whitespace_is_collapsed(self) -> None:
+        result = score(professional_summary="  Python backend\n in Engineering.  ")
+        assert result.professional_summary == "Python backend in Engineering."
+
+    def test_over_length_professional_summary_rejected(self) -> None:
+        with pytest.raises(ValidationError):
+            score(professional_summary="x" * 501)
+
 
 class TestCritique:
     def test_whitespace_is_collapsed(self) -> None:

tests/unit/test_prompts.py

@@ -35,6 +35,12 @@ def test_system_prompt_stabilises_profile_extraction() -> None:
     assert "most recent employment entry" in SYSTEM_PROMPT
 
 
+def test_system_prompt_asks_for_job_agnostic_professional_summary() -> None:
+    assert "professional_summary" in SYSTEM_PROMPT
+    assert "Ignore the job" in SYSTEM_PROMPT
+    assert "This is not summary_critique" in SYSTEM_PROMPT
+
+
 def test_injection_text_stays_inside_resume_delimiters() -> None:
     content = build_user_content("Backend engineer", INJECTION)
     resume_block = content[1]["text"]