AI Product Engineer / Tech Lead · Feb 2024–Jan 2026
Finally
AI bookkeeping, banking infrastructure, and corporate-card underwriting
What I worked on
- Joined as the first AI Product Engineer and led a three-engineer team from prototype to production on Classify AI; processed 50K+ transactions daily and reduced manual categorization by approximately 80%.
- Built retrieval-augmented classification with LangChain, Pinecone, Elasticsearch, Redis, Celery, and Django, combining transaction history, merchant enrichment, and custom charts of accounts.
- Evolved bookkeeping from CSV upload to Plaid/Teller bank ingestion, OCR-supported statements, reconciliation, and QuickBooks push; helped reduce first-month close from 4+ months to approximately 2 weeks.
- Architected reusable Plaid/Teller infrastructure for account linking, encrypted token storage, token lifecycle management, webhooks, transaction synchronization, and normalized account data.
- Built cash-based underwriting with 90-day bank data, daily-balance reconstruction, weekly recalculation, audit history, and manual overrides; supported $3M+ in credit for 50+ companies in approximately three months.
What to discuss
These are selected public facts from my résumé. I’m happy to walk through the decisions, tradeoffs, and my specific contribution in a conversation.