SentinelCorp

India Company Risk Profile API for AI Agents.

Validate GSTIN/CIN/PAN, check SEBI debarred entities, get unified risk scores — all via a free, agent-native API. 1,000 requests, no signup.

Part of the Sentinel Series alongside SentinelX402 (threat intelligence).

---

What It Does

Capability	Endpoint	What You Get
GSTIN validation	/api/v1/validate/gstin	Format + checksum + state + PAN extraction (<20ms)
CIN validation	/api/v1/validate/cin	Listing status, state, year, industry, ownership type
PAN validation	/api/v1/validate/pan	Format + entity type (Person/Company/HUF/Trust...)
Company risk profile	/api/v1/company/profile	Unified 0-100 risk score from all signals
Batch profiling	/api/v1/company/batch	Up to 100 companies in one request
SEBI debarred search	/api/v1/debarred/search	14,000+ debarred entities (NSE/BSE/SEBI)

---

Why This Exists

Indian fintech, banking, audit firms, and compliance consultants waste hours per day looking up:

• Is this GSTIN valid?
• Is this company on a SEBI defaulter list?
• What's the risk profile of this entity?

AI agents can now do this in one API call.

Example:

curl "https://YOUR_API/api/v1/company/profile?identifier=Sahara+India"

Returns:

{
  "query": "Sahara India",
  "overall_risk_score": 66.5,
  "risk_level": "high",
  "is_debarred": true,
  "debarred_matches": [
    {"matched_name": "M/s Sahara India (and its constituent partners)", "confidence": 1.0, "source": "opensanctions:nse"}
  ],
  "signals": [{"source": "sebi", "signal_type": "sebi_debarred", "severity": "critical"}]
}

---

Quickstart

curl

# Validate a GSTIN
curl "https://YOUR_API/api/v1/validate/gstin?gstin=27AAACT1234A1Z5"

# Get company risk profile (auto-detects CIN/GSTIN/PAN/name)
curl "https://YOUR_API/api/v1/company/profile?identifier=L17110MH1973PLC019786"

# Search SEBI debarred list
curl "https://YOUR_API/api/v1/debarred/search?name=sahara"

# Batch 100 companies
curl -X POST "https://YOUR_API/api/v1/company/batch" \
  -H "Content-Type: application/json" \
  -d '{"identifiers": ["Sahara", "Tata", "Reliance"]}'

Python SDK

pip install sentinelcorp

from sentinelcorp import SentinelCorp

client = SentinelCorp()
profile = client.profile("Sahara India")
print(profile["overall_risk_score"])  # 66.5
print(profile["is_debarred"])          # True

LangChain

from integrations.langchain_tool import get_sentinelcorp_tools

tools = get_sentinelcorp_tools()
agent = initialize_agent(tools, llm)
agent.run("Is Sahara India a high-risk company for us to partner with?")

MCP (Claude, Cursor)

{
  "mcpServers": {
    "sentinelcorp": {
      "command": "python",
      "args": ["integrations/mcp_server.py"]
    }
  }
}

---

Use Cases

1. KYB Automation (Know Your Business)

Before onboarding a vendor/customer, check their Indian identifiers and sanction status in one call.

2. Due Diligence at Scale

Batch-check 100 companies against SEBI debarred list in a single request.

3. Fraud Detection

Validate GSTINs on incoming invoices — catch fake GSTINs instantly with checksum verification.

4. Compliance Monitoring

Daily cron job checking your vendor list against updated SEBI debarred entities.

5. AI Agent Due Diligence

Your AI compliance agent can now verify Indian counterparties autonomously.

---

Data Sources

Source	Coverage	Legal Status
SEBI/NSE/BSE debarred	14,858 entities	Public (via OpenSanctions)
GSTIN validation	All Indian GSTINs	Pure algorithmic (Luhn mod-36)
CIN validation	All Indian companies	Pure algorithmic + metadata extraction
PAN validation	All PANs	Pure algorithmic

Future additions (with proper licensing):

• MCA V3 lookup (via Surepass/APIclub)
• Court records (via IndianKanoon API + openjustice-in/ecourts)
• GSTIN live lookup (via Jamku/KnowYourGST)
• Udyam MSME registry
• IBBI/NCLT bankruptcy orders

---

Architecture

sentinelcorp/
├── app/
│   ├── main.py                  # FastAPI app + middleware
│   ├── config.py                # Environment configuration
│   ├── database.py              # Async SQLAlchemy
│   ├── models/                  # DB models
│   ├── schemas/                 # Pydantic schemas
│   ├── routers/                 # API endpoints
│   ├── services/                # Business logic
│   │   ├── validators.py        # GSTIN/CIN/PAN (pure algorithmic)
│   │   ├── risk_scoring.py      # Noisy-OR signal combination
│   │   └── risk_service.py      # Orchestration
│   ├── scrapers/                # Data ingestion
│   │   └── sebi_defaulters.py   # OpenSanctions NSE data
│   └── data/                    # Seed scripts
├── sdk/sentinelcorp/            # Python SDK (sync + async)
├── integrations/                # Agent framework integrations
│   ├── langchain_tool.py
│   ├── crewai_tool.py
│   ├── openai_functions.py
│   └── mcp_server.py
├── tests/                       # 25 tests
├── Dockerfile                   # Production container
└── mcp.json                     # Agent discovery metadata

---

Tech Stack

• Backend: FastAPI + Uvicorn/Gunicorn
• Database: SQLite (dev) / PostgreSQL (prod)
• Validation: Pure Python checksums (no external calls)
• Data sources: OpenSanctions (NSE debarred), scrapers for public gov data
• Integrations: LangChain, CrewAI, OpenAI, MCP, Python SDK

---

Free Tier

• 1,000 requests per client — no signup, no API key
• 60 req/min rate limit per endpoint
• Batch: 100 identifiers per request
• After free tier: register for free API key with 10K/month quota

---

License

AGPL-3.0. The reference risk scoring implementation is public; the production-tuned weights, additional data sources, and proprietary integrations remain on the hosted API.

---

Related Products

• SentinelX402 — Threat intelligence API for AI agents, with CERT-In India advisories

Use both together for complete India due diligence coverage (threats + company risk).