Software Engineer, AI Capture

Notion Notion · Enterprise · San Francisco, CA · Engineering

Software Engineer, AI Capture at Notion, focusing on building advanced AI Meeting Notes and broader AI data capture features to turn conversations into durable context, tasks, and knowledge. The role involves end-to-end product ownership, improving summary quality, building agentic workflows, and ensuring enterprise readiness. Requires strong full-stack engineering skills and experience operating production systems, with a focus on AI/LLM product development and real-time media pipelines as nice-to-haves.

What you'd actually do

  1. Ship end-to-end product experiences across capture → transcript → summary → follow-ups (full-stack ownership).
  2. Make meeting & data capture feel effortless and magical (e.g., speaker identification via audio waveforms, richer in-meeting UX, smarter organization).
  3. Improve summary quality that teams trust: structure, factuality, and citations that make downstream agents and humans more capable.
  4. Raise the bar on reliability & observability across the pipeline (SLOs, debugging workflows, incident response) for realtime systems.
  5. Build agentic meeting workflows that turn discussions into tasks, follow-ups, and organized knowledge — so “we talk and things get done.”

Skills

Required

  • Strong full-stack engineering skills (frontend + backend) and excitement to own user-facing product end-to-end.
  • Product-minded craftsmanship: you sweat details, iterate quickly, and use data and user feedback to guide decisions.
  • You don’t need to be an AI expert, but you’re curious and willing to adopt AI tools to work smarter and deliver better results.

Nice to have

  • Experience with LLM / applied AI product development (prompting, evals, model integrations, or quality measurement).
  • Experience with media / realtime pipelines (audio, transcription, diarization, streaming, low-latency processing).
  • Experience building for enterprise customers (permissions models, compliance, scale, and security).

What the JD emphasized

  • 10+ years shipping production software, with a strong track record of owning features end-to-end
  • Experience building and operating production systems (debugging, on-call/incident response, performance, and reliability).
  • Ability to work across ambiguous problem spaces, align stakeholders, and drive execution with high ownership.

Other signals

  • Build agentic meeting workflows that turn discussions into tasks, follow-ups, and organized knowledge
  • Ship end-to-end product experiences across capture → transcript → summary → follow-ups (full-stack ownership)
  • Improve summary quality that teams trust: structure, factuality, and citations that make downstream agents and humans more capable.