Project 3: Auto-CRM Webhook Injection Service

Author: Shashank Jha
Alignment: Realates AI Systems (Core Service #3: Auto-CRM Injection Systems)

---

1. Business Problem

When client acquisition funnels capture hundreds of leads across WhatsApp, web forms, and voice calls, data fragmentation becomes a major liability. Furthermore, third-party CRM APIs (like GoHighLevel or HubSpot) frequently experience network drops or rate limits. If a lead webhook fails without retry logic, that client prospect is lost.

2. Solution

The auto-crm-webhook-injector microservice solves data leakage through defensive integration engineering. It transforms raw inbound lead data into standardized GoHighLevel contact schemas, generates HMAC SHA256 security headers, and executes REST API injection. Crucially, it incorporates Tenacity exponential backoff retries (up to 5 attempts) to target reliable payload delivery.

---

3. Architecture Diagram

sequenceDiagram
    autonumber
    participant AI as AI Qualification Agent
    participant Inj as AutoCRMInjector Service
    participant Sec as HMAC SHA256 Signer
    participant CRM as GoHighLevel REST API

    AI->>Inj: Dispatch Qualified Lead Object
    Inj->>Inj: Transform Schema (Split Name, Map Tags)
    Inj->>Sec: Sign Payload String
    Sec-->>Inj: X-Realates-Signature Header
    Inj->>CRM: POST /v1/contacts/ (JSON Payload)
    alt Network OK (200 OK)
        CRM-->>Inj: 200 Created + Contact ID
        Inj-->>AI: Confirmation Status
    else API Timeout / Rate Limit (503 / 429)
        CRM-->>Inj: RequestException
        Inj->>Inj: Tenacity Retry Backoff (Wait 2s, 4s...)
        Inj->>CRM: Re-attempt POST /v1/contacts/
        CRM-->>Inj: 200 Created
    end

Loading

---

4. Tech Stack

• HTTP Client: Requests
• Resiliency: Tenacity (Exponential backoff decorators)
• Security: HMAC SHA256
• Validation: Pydantic v2

---

5. Repository Structure

3-auto-crm-webhook-injector/
├── Dockerfile                 # Container manifest
├── docker-compose.yml         # Container orchestration
├── requirements.txt           # Dependency manifest
├── .env.example               # Config template
├── README.md                  # Operating documentation
├── ARCHITECTURE.md            # Architecture specs
├── DEPLOYMENT.md              # Cloud deployment instructions
├── tests/
│   └── test_injector.py       # Verification test suite
└── src/
    ├── __init__.py
    ├── crm_injector.py
    └── main.py

---

6. Setup Guide

cd portfolio/3-auto-crm-webhook-injector
cp .env.example .env
pytest tests/
python -m src.main

---

7. Tenacity Retry Implementation

@retry(
    stop=stop_after_attempt(5),
    wait=wait_exponential(multiplier=1, min=2, max=10),
    retry=retry_if_exception_type(requests.RequestException),
    reraise=True
)
def inject_contact_into_crm(self, lead_data: Dict[str, Any]) -> Dict[str, Any]:
    ...