Frontier AI Engineer

Postman is the world’s leading API platform, used by more than 45 million+ developers and 500,000 organizations, including 98% of the Fortune 500. Postman is helping developers and professionals across the globe build the API-first world by simplifying each step of the API lifecycle and streamlining collaboration—enabling users to create better APIs, faster. The company is headquartered in San Francisco and has offices in Boston, New York, Austin, Tokyo, London, and Bangalore - where Postman was founded.

Postman is privately held, with funding from Battery Ventures, BOND, Coatue, CRV, Insight Partners, and Nexus Venture Partners. Learn more at  or connect with Postman on X (Use the "Apply for this Job" box below).:
We highly recommend reading The "API-First World" graphic novel to understand the bigger picture and our vision at Postman.

We are building a small, exceptional Frontier AI Engineering team  idea is pretty simple. Some of the most important problems in AI are not solved in a lab or in a planning meeting. They are solved when great engineers sit with customers, understand what they are trying to build, see where things break, and help make it work.

Frontier AI Engineers will work directly with some of our most ambitious customers to design, build, and bring real AI and agentic systems into production. This is a hands‑on engineering role. You will write code, build prototypes, debug systems, make architecture decisions, and work side by side with customer engineering teams to get things running. You might be working on an agent that needs reliable tool use, an MCP based integration, a RAG pipeline that needs better retrieval quality, an evaluation framework for production agents, model routing across multiple LLM providers, or the identity and policy layer required to safely let agents take actions.

The work spans the full AI stack, from models and inference to agents, APIs, tools, context, evaluation, security, observability, and production infrastructure. Just as importantly, you will bring what you learn back into Postman. We do not want to build a collection of one off solutions. We want to understand the hard AI engineering problems our customers are running into, recognize the patterns that repeat, and turn those lessons into product capabilities that can help thousands of other developers.

Build real AI systems. Work with LLMs, agents, MCP, APIs, tool calling, RAG, vector search, structured outputs, multimodal models, and production AI workflows. Design agentic systems. Build agents that can reason across tools, APIs, enterprise data, and workflows while maintaining the right controls around identity, permissions, policy, and execution. Work across the model layer. Experiment with leading foundation models, model routing, inference, prompt and context engineering, fine tuning where it makes sense, and the tradeoffs between quality, latency, cost, and reliability.

Make AI measurable. Build evaluation systems for agents and LLM applications, including task success, tool correctness, hallucination detection, regression testing, safety, latency, and cost. Make AI production ready. Work on observability, tracing, guardrails, fallback strategies, rate limits, retries, caching, model gateways, and the infrastructure needed to operate AI systems reliably ve the context problem. Help customers reason about retrieval, memory, grounding, context windows, embeddings, vector databases, enterprise data, and how the right context gets to the right model or agent at the right time.

Secure agent actions. Work on authentication, authorization, secrets, API access, policy enforcement, human approval, auditability, and the controls required when AI systems…