LaunchDarkly helps teams safely manage code and AI agents in production with feature flags, progressive delivery, automated rollback, and runtime control. Backed by Bessemer, Insight and 500 Global.
About the role
As a Full Stack Engineer on LaunchDarkly’s AgentControl team, you'll build critical systems such as the web application that customers interact with on a daily basis and distributed systems that power AI evaluations at scale. These systems enable customers to control, monitor, and optimize their agent’s functionality. You’ll primarily work in Go, Python, and TypeScript and will work directly with other technologies and tools such as AWS, CockroachDB, Datadog, DeepEval, and FastAPI.
What they're looking for
- 5+ years of professional software engineering experience, with a track record of shipping production-quality full-stack features
- Hands-on experience building GenAI features, Agents, and/or working with Large Language Models (LLMs), either through direct model integration or leveraging AI APIs like OpenAI, Google Cloud AI, or AWS Bedrock
- Understanding of prompt engineering, fine-tuning models, or deploying AI services in production environments
- Experience with designing, implementing, and maintaining RESTful APIs
- Experience writing production-ready code with emphasis on quality and maintainability
- Experience with distributed systems or data ingestion
More about this role
As a Full Stack Engineer on LaunchDarkly’s AgentControl team, you'll build critical systems such as the web application that customers interact with on a daily basis and distributed systems that power AI evaluations at scale. These systems enable customers to control, monitor, and optimize their agent’s functionality. You’ll primarily work in Go, Python, and TypeScript and will work directly with other technologies and tools such as AWS, CockroachDB, Datadog, DeepEval, and FastAPI.
LaunchDarkly’s AgentControl team is on a mission to manage the complete software development lifecycle for shipping agents to production, from configuring to benchmarking to observing and beyond. AgentControl enables teams to adjust agent behavior without needing a redeploy or code change, while also running experiments, monitoring agents in production, and putting guardrails around AI as usage scales. The product spans offline evals, real-time observability, prompt and workflow optimization, and decision-making around quality, accuracy, relevance, and cost. The AgentControl team has a huge opportunity ahead of it; it's a strategic bet for LaunchDarkly, and we need your help writing the next chapter in...
Browse similar: Startup jobs · Remote jobs