Lead Product Manager, Agent Platform

Pylon•New York, NY

About The Position

Pylon is hiring a Lead Product Manager to own its Agent Platform: the foundation for Pylon’s internal agents and the way customers’ external agents work with Pylon. The role combines internal platform development with external developer capabilities. Internally, the PM will build tools and infrastructure that engineering, operations, and business teams use to build, test, ship, and monitor agents across mortgage workflows. Externally, the PM will make it straightforward and safe for customers to connect their agents through MCP servers, agent-oriented APIs, permissions, and guardrails. The role sets the direction for agents across Pylon and owns the roadmap for platform capabilities including tools, orchestration, context and memory, retrieval, evaluations, observability, and guardrails. The Lead PM will accelerate internal teams’ ability to take agents from idea to production, establish shared development patterns, and create robust evaluation practices with datasets, benchmarks, regression checks, and human review for regulated decisions. The role also ensures agent actions are traceable and explainable, makes build-versus-buy decisions, and balances cost, latency, and quality. This is a hands-on product role that includes writing and shipping code, committing to the production codebase, and building agents on the platform. The Lead PM will work with go-to-market and account management teams to help customers integrate agents, support deals, and drive customer success, while helping hire and grow the team as the platform scales. The person in this role is expected to become a trusted mortgage expert over time. Candidates must live in or be willing to relocate to Palo Alto, California, or New York City.

Requirements

At least 4 years of product management experience; an engineering background is a plus.
Experience owning and scaling a developer platform, infrastructure product, or API used by multiple teams or external customers.
Experience shipping LLM-based agents or agent tooling to production, with knowledge of tool use, MCP, multi-agent orchestration, and structured outputs.
Knowledge of evaluations, including golden datasets, LLM-as-judge, and regression testing, and ability to connect model quality to business results.
Understanding of retrieval, including RAG, embeddings, vector databases, chunking, hybrid search, and reranking.
Understanding of tradeoffs among prompt and context engineering, model choice, and fine-tuning, including effects on cost, latency, and accuracy.
Familiarity with tracing, guardrails, prompt injection, and protecting sensitive information such as borrower PII.
Ability to write code, commit to a production codebase, build prototypes, and discuss architecture in depth with engineers.
Experience building 0-to-1 enterprise software in a regulated industry.
Ability to write SQL and define metrics to assess agent performance and platform usage.
Must live in or be willing to relocate to Palo Alto, California, or New York City.
Willingness to develop deep mortgage expertise over time.

Nice To Haves

Engineering background.
Mortgage or fintech infrastructure experience.
Experience with document understanding, such as OCR, extraction, or classification.

Responsibilities

Set the direction for agents across Pylon and align product, engineering, and operations on the highest-impact opportunities.
Own the roadmap for platform capabilities, including tools, orchestration, context and memory, retrieval, evaluations, observability, and guardrails.
Help internal teams move faster from agent ideas to production and establish shared patterns to avoid duplicative work.
Develop evaluation capabilities, including shared datasets, benchmarks, regression checks, and human review for regulated decisions.
Own how external agents connect and work with Pylon, including MCP servers, APIs, scoped permissions, rate limits, and documentation for people and LLMs.
Ensure agent actions can be traced and explained, with regulatory needs addressed from the outset.
Make build-versus-buy decisions for models, vector databases, frameworks, and vendors, considering cost, latency, and quality.
Write and ship code, commit to the production codebase, and build agents on the platform.
Work with go-to-market and account management teams to integrate customer agents with Pylon, support deals, and help customers succeed.
Help hire and grow the team as the platform scales.
Become a trusted mortgage expert on the team over time.