Software Engineer, AI Application
Plane · United States
About this role
**Software Engineer, AI Application** **About Plane** Plane is building the infrastructure for how the world's work gets done. We bring projects, knowledge, and requests into one open, adaptable platform used by 55,000+ developers and teams globally. We're backed by top investors and operate across San Francisco, London, and Hyderabad. **The Role** You'll build the agent layer that lets AI participate directly in work planning, tracking, and shipping on Plane. This is product engineering—you'll write Python, design retrieval and tool surfaces, and own what you ship with real customers. **What You'll Do** • Build agent workflows with planning loops, tool calling, multi-step execution, and failure recovery • Design and ship MCP servers exposing Plane capabilities as safe, schema-defined tools • Own retrieval end-to-end: chunking, embeddings, vector storage, hybrid search, reranking • Build agent interfaces with streaming output, step visibility, approval gates, and clean interruption • Establish evaluation harnesses with datasets, regression suites, and offline/online metrics • Instrument tracing, token cost, and latency; drive down production failure modes • Design permission and audit models ensuring agents respect user access boundaries • Evaluate models, frameworks, and techniques; make clear adoption recommendations • Partner across teams to ship prototypes to production **What You'll Bring** • 2+ years shipping production software, including 1+ year with LLM-backed features • Strong Python and production service operations experience • Agent systems experience: orchestration, failure handling, loop control • Working knowledge of retrieval systems: embeddings, chunking, vector databases • Experience building evaluation harnesses for non-deterministic systems • Shipped AI features to real users (not just demos) **Nice to Have** • Built or published an MCP server • Run agents against production data with real permission models • Developer tools, project management, or complex domain model experience • Open source contributions **Tech Stack** Python, model APIs, MCP, agent orchestration, vector stores, tracing, and evaluation tooling. **Why Plane?** Real scope serving Fortune 500s and startups simultaneously. Build in the open where issues and roadmaps are public. Work on the human-and-agents future in production, not research. Your name is on the work.
Listing freshness
CronJobs last confirmed this listing 22h ago. If its source stops confirming the opening for seven days, this page is removed from active inventory.