I take on a small number of consulting engagements on production AI systems — the same work I do every day at Samaya AI, where I build the RAG and agent infrastructure behind research teams at Morgan Stanley, BlackRock, Citi, and Point72.
If your AI product works in the demo but struggles in production — reliability, latency, cost, or all three — that is the problem I solve.
Services
AI architecture audit
A review of your RAG or agent system end to end: architecture, data pipeline, retrieval, orchestration, and serving. I identify reliability, latency, security, and cost problems, then deliver a prioritized implementation roadmap your team can execute immediately.
Production AI evaluation setup
I build the evaluation backbone for an existing AI product: a test dataset that reflects real usage, an automated eval framework, observability, and a regression process — so every model, prompt, or retrieval change is measured before it ships.
AI performance sprint
A focused sprint to reduce inference latency and cost. At Samaya I cut time-to-first-token by 90% (100s → 10s) while scaling from 10 to 10,000 daily active users; I apply the same playbook to your stack.
Fractional AI infrastructure lead
Ongoing advisory for startups: several hours per week on architecture decisions, hiring, and production readiness — staff-level AI infrastructure experience in your corner without the full-time hire.
Working together
Email a short note about your system and what’s hurting. I’ll reply within two business days with an honest read on whether I can help — and if I’m not the right fit, I’ll say so.