All work
AI Implementation · Finance2026
DocSense
LLM-powered intake that extracts policy data from PDFs and emails into the CRM.
Client: Insurance broker
DocSense
PDF
PDF
PDF
AI
Policy no.97%
Premium99%
Expiry94%
The challenge
Staff manually re-keyed data from hundreds of policy documents a week, with frequent errors downstream.
What we built
A document pipeline using Claude for structured extraction, confidence scoring, a human-review queue for low-confidence fields and direct CRM sync.
Results
- ~35 staff-hours saved per week
- Extraction accuracy 97%+ after tuning
- Payback in under 3 months
Stack
- Python
- Claude API
- FastAPI
- PostgreSQL
- n8n
Ready to talk through your project?
Book a free 30-minute consultation. We will tell you honestly what it will take.