Law Firm Automated Client Intake

How It Works

1. 1
   Monitor HubSpot Webhooks

   Listen for new contact creation events from HubSpot webhooks, capturing inquiry details including case description, legal issues, client contact information, and urgency indicators

   HubSpot Webhooks API
2. 2
   Extract Case Details

   Parse incoming contact record to extract relevant fields including case_description, legal_area, opposing_party, damages_claimed, and statute_of_limitations_concerns from HubSpot custom properties

   HubSpot Contacts API
3. 3
   Analyze Case Merit

   Submit structured case information to Claude AI with firm-specific prompts to evaluate practice area fit, case strength, potential damages, and resource requirements against predefined acceptance criteria

   Claude API
4. 4
   Generate Screening Decision

   Process Claude's analysis to assign qualification scores (1-10), identify required follow-up actions, flag high-priority cases, and create structured recommendations for attorney review

   Claude API
5. 5
   Update Client Record

   Write screening results back to HubSpot including qualification_score, practice_area_match, merit_assessment, and next_action_required fields, then trigger workflow for qualified cases

   HubSpot Contacts API

Implementation

# Intake Screener Implementation

## Project Structure
```
intake_screener/
├── app.py
├── config/
│   ├── settings.py
│   └── criteria.yaml
├── handlers/
│   ├── hubspot_handler.py
│   ├── claude_handler.py
│   └── screening_engine.py
├── templates/
│   └── screening_prompts.py
└── requirements.txt
```

## Environment Variables (.env)
```
HUBSPOT_API_KEY=your_hubspot_key
CLAUDE_API_KEY=your_claude_key
WEBHOOK_SECRET=your_webhook_secret
FLASK_SECRET_KEY=random_secret
```

## Main Application (app.py)
```python
from flask import Flask, request, jsonify
from handlers.hubspot_handler import HubSpotClient
from handlers.claude_handler import ClaudeAnalyzer
from handlers.screening_engine import ScreeningEngine

app = Flask(__name__)
hubspot = HubSpotClient()
claude = ClaudeAnalyzer()
screener = ScreeningEngine(hubspot, claude)

@app.route('/webhook/hubspot', methods=['POST'])
def hubspot_webhook():
    data = request.json
    if data.get('subscriptionType') == 'contact.creation':
        contact_id = data['objectId']
        return screener.process_new_inquiry(contact_id)
    return jsonify({'status': 'ignored'})

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=5000)
```

## Deployment
1. Install dependencies: `pip install -r requirements.txt`
2. Configure HubSpot webhook pointing to `/webhook/hubspot`
3. Set up criteria.yaml with practice area definitions
4. Deploy with gunicorn: `gunicorn -w 4 -b 0.0.0.0:5000 app:app`
5. Configure reverse proxy (nginx) for SSL termination

Data Flow

Prerequisites

• -HubSpot webhook endpoint configured for contact creation events
• -Claude AI API key with sufficient usage credits
• -Custom HubSpot properties defined for screening results
• -Practice area criteria and acceptance guidelines documented
• -SSL certificate for secure webhook reception

Error Handling