Senior Machine Learning Engineer

Description

The technical stack centers on Bedrock AgentCore for secure, scalable agent deployment, LangGraph for graph‑based orchestration, and Strands Agents for structured reasoning and tool use.

• Design and ship production agents for permit workflows using LangGraph and Strands.
• Operationalize agents on Bedrock AgentCore with Runtime, Gateway, Memory, Identity, and observability.
• Build evaluation and safety harnesses with task‑level and end‑to‑end metrics.
• Integrate retrieval‑augmented generation and vector stores to provide reliable knowledge interfaces.
• Collaborate with Product/Design to create agent‑first UX in LiteTable.
• Ensure quality and reliability with fault‑tolerant flows, dashboards, and on‑call runbooks.
• Provide technical leadership by creating libraries, mentoring engineers, and reviewing designs.
• Deploy a production agent behind a feature flag with observability and basic evaluation within 30 days.
• Launch cooperating agents (intake, doc QA, code lookup) with guardrails and human‑hand‑off within 60 days.
• Contribute to a plan achieving ≥25% faster reviews, maintain on‑call reliability, and establish reusable agent templates by 90 days.
• Shape the data platform and labeling strategy for future hires.
• Be an early technical leader on a high‑impact AI product for building compliance.
• Work directly with architects, engineers, and domain experts.
• Contribute to a more efficient, sustainable, and safe built environment.
• Solve a large‑market, under‑explored AI problem.
• Own it: treat the data layer as your own responsibility.
• Solve real problems: ensure every schema and pipeline delivers ML‑ready data.
• Bi‑annual all‑hands meetings.
• Regular team‑building events.

Requirements

• Deep experience with LangGraph and Strands agent frameworks.
• 7–12+ years building high‑reliability backend or ML systems (Python/TypeScript, AWS, containers, CI/CD, IaC).
• Hands‑on expertise in RAG and vector databases (e.g., Pinecone, S3 Vectors).
• Ability to design realistic evaluation testbeds and understand agent benchmarks.
• Knowledge of safety and governance practices, including guardrails and human‑in‑the‑loop controls.
• Curiosity for building codes and experience handling PDFs/CAD in plan‑review workflows.
• Familiarity with Bedrock AgentCore components (Runtime, Memory, Gateway, Identity, Observability).
• Nice‑to‑have experience with CrewAI, AutoGen, LlamaIndex, Apryse PDF tools, geospatial data, or AEC/reg‑tech.

Benefits

• Competitive base salary and employee equity program.
• Performance‑based annual bonuses.
• Premium health, dental, and vision coverage.
• 401(k) retirement plan.
• Generous parental leave.
• Monthly wellness stipend and mental‑health support.
• Weekly catered team lunches in NYC office.
• Unlimited PTO.
• Hybrid work: 4 days in office, 3 days during summer.

Training + Development

Interview process

Visa Sponsorship

Security clearance