Production LLM systems
Agents, search, and document intelligence that stay up, stay observable, and stay inside the constraints of a real financial workflow.
Principal AI Engineer · CFA Charterholder
AI Engineer
Mehdi Zare, CFA is an AI engineer who ships production systems in finance — from prototype to something operations can run.
What this AI engineer does
Systems that have to run in finance, not just look good in a demo
Domain-first architecture
I learn the market, risk, or operations problem before I pick a model. The system has to survive review by people who own P&L and audit.
Delivery past the demo
Most AI work dies between a working prototype and a Monday-morning product. I take ownership of that gap — evaluation, monitoring, and handoff.
Finance is the domain
The CFA is not decoration — it is how the work stays honest
Plenty of teams can stand up a chatbot. Fewer can put a model next to a funding forecast, a research workflow, or a risk process and have it survive contact with operations.
I trained as a physicist, then earned an MBA in finance and the CFA charter. That combination is the job: learn the domain first, then engineer the system that ships.
Selected finance work
Proof that the AI engineer title is earned in markets, not slides
Quantitative Analysis Manager
Capital One
Led machine-learning liquidity forecasting across finance, data engineering, and compliance — the kind of model work that has to reconcile with how a bank actually funds itself.
CFA Institute
CFA Charter
The charter is how I talk risk, valuation, and capital markets with the people who will live with the system — not just the team that built the notebook.
Co-Founder
Adviser
Co-built a virtual investment adviser that turned conversation into personalized financial visualizations people could use, not just admire.
Senior Consultant
FI Consulting
Earned the CFA while contributing financial data work with the Office of Financial Research — translating market and policy context into usable analytical output.
Need an AI engineer on a live problem?
Working on an AI initiative that needs to ship?
Let's Talk
Prefer email? Start a conversation.