docintel-platform 1.0.2__tar.gz → 1.2.0__tar.gz

docintel_platform-1.2.0/PKG-INFO

@@ -0,0 +1,222 @@
+Metadata-Version: 2.4
+Name: docintel-platform
+Version: 1.2.0
+Summary: Enterprise document intelligence API for PDF compliance, multi-format extraction, structuring, and summarization.
+Author: Babandeep Singh
+License-Expression: MIT
+Project-URL: Homepage, https://github.com/baban9/document-intelligence-platform
+Project-URL: Repository, https://github.com/baban9/document-intelligence-platform
+Project-URL: Documentation, https://github.com/baban9/document-intelligence-platform#readme
+Project-URL: Issues, https://github.com/baban9/document-intelligence-platform/issues
+Keywords: nlp,pdf,flask,document-ai,ocr,pii,presidio,openapi,document-intelligence,compliance
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.9
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Framework :: Flask
+Classifier: Topic :: Text Processing
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Classifier: Typing :: Typed
+Requires-Python: >=3.9
+Description-Content-Type: text/markdown
+Requires-Dist: flask>=3.0.3
+Requires-Dist: werkzeug>=3.0.3
+Requires-Dist: pymupdf>=1.24.10
+Requires-Dist: scikit-learn>=1.5.2
+Requires-Dist: networkx>=3.2.1
+Requires-Dist: numpy>=1.26.4
+Requires-Dist: gunicorn>=23.0.0
+Requires-Dist: pyyaml>=6.0.2
+Requires-Dist: requests>=2.32.3
+Requires-Dist: prometheus-client>=0.21.0
+Provides-Extra: dev
+Requires-Dist: pytest>=8.3.3; extra == "dev"
+Requires-Dist: build>=1.2.2; extra == "dev"
+Requires-Dist: twine>=5.1.1; extra == "dev"
+Requires-Dist: fakeredis>=2.26.2; extra == "dev"
+Requires-Dist: prometheus-client>=0.21.0; extra == "dev"
+Requires-Dist: python-docx>=1.1.2; extra == "dev"
+Requires-Dist: openpyxl>=3.1.5; extra == "dev"
+Provides-Extra: documents
+Requires-Dist: python-docx>=1.1.2; extra == "documents"
+Requires-Dist: openpyxl>=3.1.5; extra == "documents"
+Provides-Extra: ocr
+Requires-Dist: easyocr>=1.7.2; extra == "ocr"
+Requires-Dist: presidio-analyzer>=2.2.354; extra == "ocr"
+Requires-Dist: spacy>=3.7.0; extra == "ocr"
+Requires-Dist: opencv-python-headless>=4.10.0; extra == "ocr"
+Requires-Dist: torch>=2.4.1; extra == "ocr"
+Provides-Extra: ui
+Requires-Dist: gradio>=4.44.0; extra == "ui"
+Requires-Dist: requests>=2.32.3; extra == "ui"
+Provides-Extra: llm
+Requires-Dist: openai>=1.54.0; extra == "llm"
+Provides-Extra: jobs
+Requires-Dist: redis>=5.0.8; extra == "jobs"
+Requires-Dist: rq>=1.16.2; extra == "jobs"
+Provides-Extra: storage
+Requires-Dist: boto3>=1.35.0; extra == "storage"
+Provides-Extra: auth
+Requires-Dist: flask-limiter>=3.8.0; extra == "auth"
+Requires-Dist: PyJWT>=2.9.0; extra == "auth"
+Requires-Dist: cryptography>=43.0.0; extra == "auth"
+Provides-Extra: all
+Requires-Dist: easyocr>=1.7.2; extra == "all"
+Requires-Dist: presidio-analyzer>=2.2.354; extra == "all"
+Requires-Dist: spacy>=3.7.0; extra == "all"
+Requires-Dist: opencv-python-headless>=4.10.0; extra == "all"
+Requires-Dist: torch>=2.4.1; extra == "all"
+Requires-Dist: openai>=1.54.0; extra == "all"
+Requires-Dist: redis>=5.0.8; extra == "all"
+Requires-Dist: rq>=1.16.2; extra == "all"
+Requires-Dist: boto3>=1.35.0; extra == "all"
+Requires-Dist: prometheus-client>=0.21.0; extra == "all"
+Requires-Dist: flask-limiter>=3.8.0; extra == "all"
+Requires-Dist: PyJWT>=2.9.0; extra == "all"
+Requires-Dist: cryptography>=43.0.0; extra == "all"
+Requires-Dist: gradio>=4.44.0; extra == "all"
+Requires-Dist: python-docx>=1.1.2; extra == "all"
+Requires-Dist: openpyxl>=3.1.5; extra == "all"
+
+# Document Intelligence Platform
+
+[![Python 3.9+](https://img.shields.io/badge/python-3.9+-blue.svg)](https://www.python.org/downloads/)
+[![Flask](https://img.shields.io/badge/flask-3.0+-green.svg)](https://flask.palletsprojects.com/)
+[![Docker](https://img.shields.io/badge/docker-compose-ready-blue.svg)](docker-compose.yml)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
+
+Enterprise document intelligence API: PDF compliance (OCR, PII, redaction), LLM structuring, and multi-format text workflows (Word, Excel, CSV, plain text).
+
+**Version:** 1.2.0 | **PyPI:** [docintel-platform](https://pypi.org/project/docintel-platform/)
+
+---
+
+## Quick start
+
+**Docker (API + Gradio UI + worker):**
+
+```bash
+git clone https://github.com/baban9/document-intelligence-platform.git
+cd document-intelligence-platform
+cp .env.example .env   # optional: ports, LLM key, auth
+make docker-up
+```
+
+| Service | URL |
+|---------|-----|
+| API | http://127.0.0.1:5000 |
+| Interactive API docs | http://127.0.0.1:5000/docs |
+| Gradio UI | http://127.0.0.1:7860 |
+| Health | http://127.0.0.1:5000/health |
+
+Gradio includes a **Document process** tab (unified pipeline). It needs the API plus a Redis worker (`worker` service in compose, or `make run-worker` locally).
+
+**pip install:**
+
+```bash
+pip install docintel-platform
+pip install "docintel-platform[all]"        # OCR, LLM, jobs, auth, UI, office formats
+pip install "docintel-platform[documents]"  # Word and Excel only
+```
+
+**Python client:**
+
+```python
+from docintel import DocintelClient
+
+client = DocintelClient("http://127.0.0.1:5000", api_key="your-key")
+summary = client.summarize(report_text, sentences=3)
+report = client.process_document("policy.docx", include_pii=True)
+```
+
+---
+
+## Capabilities
+
+| Area | Endpoints | Notes |
+|------|-----------|-------|
+| PDF annotate | `POST /v1/pdf/annotate` | Regex highlight, redact, markup |
+| PDF PII scan | `POST /v1/pdf/detect-sensitive` | Presidio + OCR for scanned PDFs |
+| PDF structure | `POST /v1/pdf/structure` | OCR + LLM curated PDF (needs LLM key) |
+| Documents | `POST /v1/documents/*` | Identify, extract, classify, summarize, PII, compare, **process** |
+| Text | `POST /v1/text/summarize` | TextRank extractive summary |
+| Batch | `POST /v1/batch` | Async summarize, classify, detect_pii, process |
+| Jobs | `GET /v1/jobs/{id}` | Poll async work (`?async=true`; default in Docker when Redis is up) |
+| Ops | `GET /health`, `GET /metrics` | Health and Prometheus-friendly metrics |
+
+**Supported uploads (text workflows):** PDF, DOCX, XLSX, CSV, JSON, TXT, MD.
+
+**PDF-only routes** (annotate, sensitive, structure) return HTTP 415 for other types. Use `/v1/documents/extract-text` or `/v1/documents/process` for office files.
+
+Full request and response schemas: **http://127.0.0.1:5000/docs** (OpenAPI).
+
+---
+
+## Example requests
+
+```bash
+# Sensitive PDF (digital or scanned)
+curl -X POST http://127.0.0.1:5000/v1/pdf/detect-sensitive \
+  -F "file=@contract.pdf" -F "action=Highlight" -o marked.pdf
+
+# Unified document pipeline (extract + classify + summarize + PII)
+curl -X POST http://127.0.0.1:5000/v1/documents/process \
+  -F "file=@policy.docx" -F "sentences=3"
+
+# Async: add ?async=true, then poll /v1/jobs/<job_id>
+curl -X POST "http://127.0.0.1:5000/v1/documents/process?async=true" \
+  -F "file=@policy.docx"
+```
+
+---
+
+## Local development
+
+```bash
+make setup              # venv + dev deps
+make setup-ocr          # EasyOCR, Presidio, spaCy model
+make setup-llm          # OpenAI client (structure endpoint)
+make setup-ui           # Gradio
+make run                # API on :5000
+make run-worker         # RQ worker (separate terminal, needs Redis)
+make run-ui             # Gradio on :7860
+make test
+make eval               # offline quality report (summary, classify, process, PII)
+```
+
+Copy `.env.example` to `.env` for `DOCINTEL_LLM_API_KEY`, auth keys, Redis, and S3. See comments in that file for all variables.
+
+---
+
+## Documentation
+
+| Doc | Contents |
+|-----|----------|
+| [/docs](http://127.0.0.1:5000/docs) | Live OpenAPI / Swagger (authoritative API reference) |
+| [docs/PLATFORM.md](docs/PLATFORM.md) | Jobs, auth, storage, ops layout |
+| [docs/PRODUCTION.md](docs/PRODUCTION.md) | Checklist, latency, failure modes |
+| [docs/ROADMAP.md](docs/ROADMAP.md) | Milestones and history |
+| [docs/adr/](docs/adr/) | Architecture decision records |
+
+---
+
+## Project layout
+
+```
+src/docintel/
+  routes/          HTTP API
+  capabilities/    Compliance, extraction, understanding
+  jobs/            Async queue (Redis + RQ)
+  auth/            API keys, OIDC, rate limits
+  storage/         Local or S3 artifacts
+  ops/             Logging and metrics
+```
+
+---
+
+## License
+
+MIT. See [LICENSE](LICENSE).

docintel_platform-1.2.0/README.md

@@ -0,0 +1,139 @@
+# Document Intelligence Platform
+
+[![Python 3.9+](https://img.shields.io/badge/python-3.9+-blue.svg)](https://www.python.org/downloads/)
+[![Flask](https://img.shields.io/badge/flask-3.0+-green.svg)](https://flask.palletsprojects.com/)
+[![Docker](https://img.shields.io/badge/docker-compose-ready-blue.svg)](docker-compose.yml)
+[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE)
+
+Enterprise document intelligence API: PDF compliance (OCR, PII, redaction), LLM structuring, and multi-format text workflows (Word, Excel, CSV, plain text).
+
+**Version:** 1.2.0 | **PyPI:** [docintel-platform](https://pypi.org/project/docintel-platform/)
+
+---
+
+## Quick start
+
+**Docker (API + Gradio UI + worker):**
+
+```bash
+git clone https://github.com/baban9/document-intelligence-platform.git
+cd document-intelligence-platform
+cp .env.example .env   # optional: ports, LLM key, auth
+make docker-up
+```
+
+| Service | URL |
+|---------|-----|
+| API | http://127.0.0.1:5000 |
+| Interactive API docs | http://127.0.0.1:5000/docs |
+| Gradio UI | http://127.0.0.1:7860 |
+| Health | http://127.0.0.1:5000/health |
+
+Gradio includes a **Document process** tab (unified pipeline). It needs the API plus a Redis worker (`worker` service in compose, or `make run-worker` locally).
+
+**pip install:**
+
+```bash
+pip install docintel-platform
+pip install "docintel-platform[all]"        # OCR, LLM, jobs, auth, UI, office formats
+pip install "docintel-platform[documents]"  # Word and Excel only
+```
+
+**Python client:**
+
+```python
+from docintel import DocintelClient
+
+client = DocintelClient("http://127.0.0.1:5000", api_key="your-key")
+summary = client.summarize(report_text, sentences=3)
+report = client.process_document("policy.docx", include_pii=True)
+```
+
+---
+
+## Capabilities
+
+| Area | Endpoints | Notes |
+|------|-----------|-------|
+| PDF annotate | `POST /v1/pdf/annotate` | Regex highlight, redact, markup |
+| PDF PII scan | `POST /v1/pdf/detect-sensitive` | Presidio + OCR for scanned PDFs |
+| PDF structure | `POST /v1/pdf/structure` | OCR + LLM curated PDF (needs LLM key) |
+| Documents | `POST /v1/documents/*` | Identify, extract, classify, summarize, PII, compare, **process** |
+| Text | `POST /v1/text/summarize` | TextRank extractive summary |
+| Batch | `POST /v1/batch` | Async summarize, classify, detect_pii, process |
+| Jobs | `GET /v1/jobs/{id}` | Poll async work (`?async=true`; default in Docker when Redis is up) |
+| Ops | `GET /health`, `GET /metrics` | Health and Prometheus-friendly metrics |
+
+**Supported uploads (text workflows):** PDF, DOCX, XLSX, CSV, JSON, TXT, MD.
+
+**PDF-only routes** (annotate, sensitive, structure) return HTTP 415 for other types. Use `/v1/documents/extract-text` or `/v1/documents/process` for office files.
+
+Full request and response schemas: **http://127.0.0.1:5000/docs** (OpenAPI).
+
+---
+
+## Example requests
+
+```bash
+# Sensitive PDF (digital or scanned)
+curl -X POST http://127.0.0.1:5000/v1/pdf/detect-sensitive \
+  -F "file=@contract.pdf" -F "action=Highlight" -o marked.pdf
+
+# Unified document pipeline (extract + classify + summarize + PII)
+curl -X POST http://127.0.0.1:5000/v1/documents/process \
+  -F "file=@policy.docx" -F "sentences=3"
+
+# Async: add ?async=true, then poll /v1/jobs/<job_id>
+curl -X POST "http://127.0.0.1:5000/v1/documents/process?async=true" \
+  -F "file=@policy.docx"
+```
+
+---
+
+## Local development
+
+```bash
+make setup              # venv + dev deps
+make setup-ocr          # EasyOCR, Presidio, spaCy model
+make setup-llm          # OpenAI client (structure endpoint)
+make setup-ui           # Gradio
+make run                # API on :5000
+make run-worker         # RQ worker (separate terminal, needs Redis)
+make run-ui             # Gradio on :7860
+make test
+make eval               # offline quality report (summary, classify, process, PII)
+```
+
+Copy `.env.example` to `.env` for `DOCINTEL_LLM_API_KEY`, auth keys, Redis, and S3. See comments in that file for all variables.
+
+---
+
+## Documentation
+
+| Doc | Contents |
+|-----|----------|
+| [/docs](http://127.0.0.1:5000/docs) | Live OpenAPI / Swagger (authoritative API reference) |
+| [docs/PLATFORM.md](docs/PLATFORM.md) | Jobs, auth, storage, ops layout |
+| [docs/PRODUCTION.md](docs/PRODUCTION.md) | Checklist, latency, failure modes |
+| [docs/ROADMAP.md](docs/ROADMAP.md) | Milestones and history |
+| [docs/adr/](docs/adr/) | Architecture decision records |
+
+---
+
+## Project layout
+
+```
+src/docintel/
+  routes/          HTTP API
+  capabilities/    Compliance, extraction, understanding
+  jobs/            Async queue (Redis + RQ)
+  auth/            API keys, OIDC, rate limits
+  storage/         Local or S3 artifacts
+  ops/             Logging and metrics
+```
+
+---
+
+## License
+
+MIT. See [LICENSE](LICENSE).

{docintel_platform-1.0.2 → docintel_platform-1.2.0}/pyproject.toml

@@ -4,16 +4,16 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "docintel-platform"
-version = "1.0.2"
-description = "Document intelligence API and Python client for PDF OCR, PII detection, LLM structuring, matching, and summarization."
+version = "1.2.0"
+description = "Enterprise document intelligence API for PDF compliance, multi-format extraction, structuring, and summarization."
 readme = "README.md"
 license = "MIT"
 license-files = []
 requires-python = ">=3.9"
 authors = [{name = "Babandeep Singh"}]
 keywords = [
-    "nlp", "pdf", "flask", "document-ai", "resume-matching",
-    "ocr", "pii", "presidio", "openapi", "document-intelligence",
+    "nlp", "pdf", "flask", "document-ai",
+    "ocr", "pii", "presidio", "openapi", "document-intelligence", "compliance",
 ]
 classifiers = [
     "Development Status :: 4 - Beta",
@@ -38,6 +38,7 @@ dependencies = [
     "gunicorn>=23.0.0",
     "pyyaml>=6.0.2",
     "requests>=2.32.3",
+    "prometheus-client>=0.21.0",
 ]
 
 [project.optional-dependencies]
@@ -46,6 +47,13 @@ dev = [
     "build>=1.2.2",
     "twine>=5.1.1",
     "fakeredis>=2.26.2",
+    "prometheus-client>=0.21.0",
+    "python-docx>=1.1.2",
+    "openpyxl>=3.1.5",
+]
+documents = [
+    "python-docx>=1.1.2",
+    "openpyxl>=3.1.5",
 ]
 ocr = [
     "easyocr>=1.7.2",
@@ -65,6 +73,9 @@ jobs = [
     "redis>=5.0.8",
     "rq>=1.16.2",
 ]
+storage = [
+    "boto3>=1.35.0",
+]
 auth = [
     "flask-limiter>=3.8.0",
     "PyJWT>=2.9.0",
@@ -79,10 +90,14 @@ all = [
     "openai>=1.54.0",
     "redis>=5.0.8",
     "rq>=1.16.2",
+    "boto3>=1.35.0",
+    "prometheus-client>=0.21.0",
     "flask-limiter>=3.8.0",
     "PyJWT>=2.9.0",
     "cryptography>=43.0.0",
     "gradio>=4.44.0",
+    "python-docx>=1.1.2",
+    "openpyxl>=3.1.5",
 ]
 
 [project.scripts]

{docintel_platform-1.0.2 → docintel_platform-1.2.0}/src/docintel/__init__.py

@@ -2,5 +2,5 @@
 
 from docintel.client import DocintelClient, DocintelError
 
-__version__ = "1.0.0"
+__version__ = "1.2.0"
 __all__ = ["DocintelClient", "DocintelError", "__version__"]

{docintel_platform-1.0.2 → docintel_platform-1.2.0}/src/docintel/app.py

@@ -8,9 +8,10 @@ from docintel.auth.limiter import init_limiter
 from docintel.auth.middleware import register_auth
 from docintel.ops.logging import configure_logging
 from docintel.ops.middleware import register_request_hooks
+from docintel.routes.documents import documents_bp
+from docintel.routes.batch import batch_bp
 from docintel.routes.jobs import jobs_bp
 from docintel.routes.openapi_docs import docs_bp
-from docintel.routes.match import match_bp
 from docintel.routes.ops import ops_bp
 from docintel.routes.pdf import pdf_bp
 from docintel.routes.text import text_bp
@@ -36,9 +37,10 @@ def create_app(config: type[Config] = Config) -> Flask:
         )
 
     app.register_blueprint(docs_bp)
+    app.register_blueprint(documents_bp)
     app.register_blueprint(pdf_bp)
     app.register_blueprint(jobs_bp)
-    app.register_blueprint(match_bp)
+    app.register_blueprint(batch_bp)
     app.register_blueprint(text_bp)
     app.register_blueprint(ops_bp)
 

docintel_platform-1.2.0/src/docintel/capabilities/__init__.py

@@ -0,0 +1 @@
+"""Enterprise capability modules."""

docintel_platform-1.2.0/src/docintel/capabilities/compliance/__init__.py

@@ -0,0 +1,14 @@
+"""Compliance capabilities (PII detection, sensitive PDF scanning)."""
+
+from docintel.capabilities.compliance.pii import PIIHit, detect_pii_in_text, list_supported_entities, mask_pii_in_text
+from docintel.capabilities.compliance.presets import DEFAULT_PII_ENTITIES, MIN_NATIVE_TEXT_CHARS, OCR_RENDER_SCALE
+
+__all__ = [
+    "DEFAULT_PII_ENTITIES",
+    "MIN_NATIVE_TEXT_CHARS",
+    "OCR_RENDER_SCALE",
+    "PIIHit",
+    "detect_pii_in_text",
+    "list_supported_entities",
+    "mask_pii_in_text",
+]

{docintel_platform-1.0.2/src/docintel/services/pdf → docintel_platform-1.2.0/src/docintel/capabilities/compliance}/pii.py

@@ -6,7 +6,7 @@ from dataclasses import dataclass
 from functools import lru_cache
 from typing import Sequence
 
-from docintel.services.pdf.presets import DEFAULT_PII_ENTITIES
+from docintel.capabilities.compliance.presets import DEFAULT_PII_ENTITIES
 
 
 @dataclass(frozen=True)
@@ -89,7 +89,9 @@ def mask_pii_in_text(
 
     Returns masked text and the number of entities redacted.
     """
-    hits = detect_pii_in_text(
+    from docintel.services.pdf import pii as pii_compat
+
+    hits = pii_compat.detect_pii_in_text(
         text,
         entities=entities,
         language=language,

docintel_platform-1.2.0/src/docintel/capabilities/compliance/presets.py

@@ -0,0 +1,80 @@
+"""Default Presidio entity presets (extend via API or custom recognizers)."""
+
+# Core Presidio entities suitable for legal, finance, and compliance workflows.
+DEFAULT_PII_ENTITIES: tuple[str, ...] = (
+    "EMAIL_ADDRESS",
+    "PHONE_NUMBER",
+    "US_SSN",
+    "CREDIT_CARD",
+    "US_BANK_NUMBER",
+    "US_DRIVER_LICENSE",
+    "US_ITIN",
+    "US_PASSPORT",
+    "PERSON",
+    "LOCATION",
+    "DATE_TIME",
+    "IP_ADDRESS",
+    "IBAN_CODE",
+    "MEDICAL_LICENSE",
+    "URL",
+)
+
+VERTICAL_ENTITY_PRESETS: dict[str, tuple[str, ...]] = {
+    "general": DEFAULT_PII_ENTITIES,
+    "healthcare": (
+        "PERSON",
+        "PHONE_NUMBER",
+        "EMAIL_ADDRESS",
+        "DATE_TIME",
+        "LOCATION",
+        "US_SSN",
+        "MEDICAL_LICENSE",
+        "US_DRIVER_LICENSE",
+        "URL",
+    ),
+    "financial": (
+        "PERSON",
+        "EMAIL_ADDRESS",
+        "PHONE_NUMBER",
+        "US_SSN",
+        "CREDIT_CARD",
+        "US_BANK_NUMBER",
+        "IBAN_CODE",
+        "US_ITIN",
+        "DATE_TIME",
+        "LOCATION",
+        "URL",
+    ),
+    "legal": (
+        "PERSON",
+        "EMAIL_ADDRESS",
+        "PHONE_NUMBER",
+        "LOCATION",
+        "DATE_TIME",
+        "US_PASSPORT",
+        "US_DRIVER_LICENSE",
+        "US_SSN",
+        "URL",
+    ),
+}
+
+# Minimum extracted characters before a page is treated as scanned (OCR fallback).
+MIN_NATIVE_TEXT_CHARS = 20
+
+# EasyOCR render scale (higher improves accuracy, increases memory).
+OCR_RENDER_SCALE = 2.0
+
+
+def list_vertical_presets() -> dict[str, list[str]]:
+    """Return named entity packs for vertical workflows."""
+    return {name: list(entities) for name, entities in VERTICAL_ENTITY_PRESETS.items()}
+
+
+def entities_for_vertical(name: str) -> tuple[str, ...]:
+    """Resolve a vertical preset name to a Presidio entity list."""
+    key = name.strip().lower()
+    try:
+        return VERTICAL_ENTITY_PRESETS[key]
+    except KeyError as exc:
+        valid = ", ".join(sorted(VERTICAL_ENTITY_PRESETS))
+        raise ValueError(f"Unknown vertical preset '{name}'. Choose from: {valid}") from exc

{docintel_platform-1.0.2/src/docintel/services/pdf → docintel_platform-1.2.0/src/docintel/capabilities/compliance}/sensitive.py

@@ -18,7 +18,7 @@ from docintel.services.pdf.ocr import (
     page_has_native_text,
     rects_for_char_range,
 )
-from docintel.services.pdf.pii import PIIHit, detect_pii_in_text
+from docintel.capabilities.compliance.pii import PIIHit
 from docintel.services.pdf.search import search_for_text
 
 
@@ -124,7 +124,9 @@ def detect_sensitive_pdf(
 
     Uses native PDF text when available. Falls back to EasyOCR for scanned pages.
     """
-    _ensure_ocr_stack()
+    from docintel.services.pdf import sensitive as sensitive_compat
+
+    sensitive_compat._ensure_ocr_stack()
     selected_action = action if isinstance(action, Action) else Action.from_value(action)
     if selected_action == Action.REMOVE:
         raise ValueError("Action 'Remove' is not supported for sensitive detection.")
@@ -159,7 +161,7 @@ def detect_sensitive_pdf(
             indexed = []
             page_text = page.get_text("text")
 
-        hits = detect_pii_in_text(page_text, entities=entities, min_score=min_score)
+        hits = sensitive_compat.detect_pii_in_text(page_text, entities=entities, min_score=min_score)
         if pattern:
             hits.extend(_regex_hits(page_text, pattern))
 

docintel_platform-1.2.0/src/docintel/capabilities/extraction/__init__.py

@@ -0,0 +1,29 @@
+"""Document extraction capabilities (OCR, LLM structuring)."""
+
+from docintel.capabilities.extraction.ocr import (
+    OCRSpan,
+    build_indexed_text,
+    embed_invisible_text_layer,
+    extract_page_ocr,
+    merge_rects,
+    page_has_native_text,
+    rects_for_char_range,
+)
+from docintel.capabilities.extraction.structure_schema import (
+    SectionBlock,
+    StructuredDocument,
+    StructuredPage,
+)
+
+__all__ = [
+    "OCRSpan",
+    "SectionBlock",
+    "StructuredDocument",
+    "StructuredPage",
+    "build_indexed_text",
+    "embed_invisible_text_layer",
+    "extract_page_ocr",
+    "merge_rects",
+    "page_has_native_text",
+    "rects_for_char_range",
+]

docintel_platform-1.2.0/src/docintel/capabilities/extraction/formats/__init__.py

@@ -0,0 +1,27 @@
+"""Multi-format document identification and text extraction."""
+
+from docintel.capabilities.extraction.formats.extract import extract_document_text
+from docintel.capabilities.extraction.formats.models import (
+    DocumentKind,
+    DocumentProfile,
+    ExtractionResult,
+    IdentificationResult,
+)
+from docintel.capabilities.extraction.formats.registry import (
+    get_profile,
+    list_supported_types,
+    profiles_for_kind,
+)
+from docintel.capabilities.extraction.formats.sniff import identify_document
+
+__all__ = [
+    "DocumentKind",
+    "DocumentProfile",
+    "ExtractionResult",
+    "IdentificationResult",
+    "extract_document_text",
+    "get_profile",
+    "identify_document",
+    "list_supported_types",
+    "profiles_for_kind",
+]