Senior AI Platform Architect
Late-stage productivity SaaS, millions of users
About the role
The client had shipped AI features quickly and needed someone to turn them into a platform: orchestration, evaluation, guardrails, model routing, and the backend architecture that agentic features run on under real load. This was a software engineering search, not a research one. The heaviest weight went to engineers who had designed and operated agentic systems in production and could talk about reliability, latency, and cost as fluently as model quality.
What the role owned
- Architecture for agentic AI systems in production: orchestration, context and memory, tool selection, and human-in-the-loop flows.
- Evaluation frameworks, monitoring, guardrails, model routing, and fallbacks.
- The distributed, event-driven backend that AI features run on at scale.
- Technical leadership across product engineering teams adopting the platform.
What the client was looking for
- Strong backend and distributed systems depth, with production AI systems shipped to real users.
- Hands-on with LLM APIs, queues and messaging, and cloud deployment at scale.
- Can reason about reliability, latency, and cost tradeoffs, not just model performance.
- Experience influencing architecture across multiple teams.
Nice to have
- Production use of LangGraph, LangChain, DSPy, or similar orchestration frameworks.
- Retrieval and vector database experience.
How the search ran
Confidential search with the hiring manager. First short list of three, hired from it. Closed in 47 days.
Similar roles
Senior Platform Engineer (Founding)
Venture-backed applied AI company, early stage, NYC
Senior / Staff AI Platform Engineer
Late-stage productivity SaaS, millions of users
Forward Deployed Engineer
Venture-backed applied AI company, early stage, NYC