Software Engineering Manager, AI Observability & Evals Platform (New York, NY)
About the role
About
the Team The LangSmith team owns and builds LangChain's core platform for observability, evaluation, and production reliability of AI systems. From tracing and annotation to run rules, evaluations, and beyond, this team owns LangSmith end-to-end. If you want to define what great AI observability looks like at production scale, this is where that work gets done.
About
the role We’re looking for an Engineering Manager to lead
the team building LangSmith, our observability and evaluation platform for LLM applications. You’ll define the technical direction, grow and mentor a world-class engineering team, and partner with product and design to deliver features that help developers build and deploy trustworthy AI systems with confidence.
What you'll do Build, mentor, and grow a high-performing team of engineers, fostering collaboration, ownership, and accountability Strengthen LangChain’s engineering culture through mentorship, high-quality code, and technical excellence Shape long-term technical direction and ensure scalability and reliability of LangSmith AI Observability Platform Partner with product and design teams to define scope, sequence, and success criteria for key initiatives Maintain a high bar for technical excellence while making sure
the team moves with urgency and focus Lead by example in writing clean, maintainable, and well-tested code in Go/Python and Typescript Engage directly with customers to understand their needs and translate them into actionable product improvements What you'll bring Experience managing or tech-leading teams delivering complex full-stack products Strong technical background in Go/Python and React/Typescript Solid understanding of databases (Postgres, Redis and ClickHouse) Proven ability to design high throughput data-intensive systems Bias towards shipping quickly and learning from customers Strong communication skills—able to collaborate with internal stakeholders and engage directly with customers