{
  "schema_version": "1.1",
  "id": "archive:https://www.youtube.com/watch?v=17-YSUHo6Lk",
  "slug": "agentic-sdlc-at-uber-uday-kiran-medisetty-adam-huda-uber-1ugtaxn",
  "url": "https://feed7.dev/p/agentic-sdlc-at-uber-uday-kiran-medisetty-adam-huda-uber-1ugtaxn",
  "title": "Agentic SDLC at Uber — Uday Kiran Medisetty & Adam Huda, Uber",
  "why_included": "Uber’s agent adoption rests on shared gateways, ready-to-run environments, skills, and a context graph. The operational bottleneck is shifting from code generation to validation and capacity.",
  "summary": "Uber says agents now produce **more than 70% of pull requests**, while lines of code per engineer doubled year over year. It also reports **250+ automated migrations** covering **9 million lines**, backed by gateways, cloud environments, skills, a context graph, and an internal assistant.",
  "practical_implication": "Builders scaling coding agents should standardize model and tool access, pre-provision isolated workspaces, move validation into the inner loop, and manage recurring maintenance centrally. Feed review outcomes back into skills instead of launching unbounded autonomous loops.",
  "agent_context": "Uber says agents now produce **more than 70% of pull requests**, while lines of code per engineer doubled year over year. It also reports **250+ automated migrations** covering **9 million lines**, backed by gateways, cloud environments, skills, a context graph, and an internal assistant.\n\nBuilders scaling coding agents should standardize model and tool access, pre-provision isolated workspaces, move validation into the inner loop, and manage recurring maintenance centrally. Feed review outcomes back into skills instead of launching unbounded autonomous loops.\n\nThe building blocks are in different stages of maturity and rollout, and lines changed do not measure product value or correctness. Uber also identifies CI capacity, experiment capacity, and decision-making as emerging constraints.",
  "source": {
    "name": "AI Engineer",
    "url": "https://www.youtube.com/watch?v=17-YSUHo6Lk",
    "published_at": "2026-08-21T13:00:06.000Z"
  },
  "source_class": "video",
  "content_type": "Video",
  "layer": "industry",
  "domains": [
    "coding"
  ],
  "topics": [
    "coding-agents",
    "adoption",
    "enterprise"
  ],
  "verification": {
    "status": "source_linked",
    "label": "Source Linked",
    "method": "source_feed",
    "verified_at": null
  },
  "uncertainty": [
    "The building blocks are in different stages of maturity and rollout, and lines changed do not measure product value or correctness. Uber also identifies CI capacity, experiment capacity, and decision-making as emerging constraints."
  ],
  "connected_context": {
    "meaning": "Uber turns the broad enterprise shift from assistance to execution into a large-scale operating case with unusually concrete adoption and migration figures. More importantly, it identifies the infrastructure and organizational constraints behind that scale: governed access, isolated environments, inner-loop validation, maintained skills, shared context, and centralized ownership. The evidence confirms throughput and reach, not product value or correctness.",
    "corpus_size": 545,
    "generated_at": "2026-08-23T18:05:29.043Z",
    "connections": [
      {
        "title": "From assistance to execution: How enterprises put AI to work",
        "source_name": "OpenAI",
        "source_url": "https://openai.com/index/how-enterprises-put-ai-to-work",
        "feed7_url": "https://feed7.dev/p/how-enterprises-put-ai-to-work-0p0rqih",
        "reason": "Uber supplies concrete scale, architecture, and operating constraints for the broader claimed shift from AI assistance toward agent execution."
      },
      {
        "title": "The Dirty Secret of Forward Deployed Engineering — Natalie Meurer, Sierra",
        "source_name": "AI Engineer",
        "source_url": "https://www.youtube.com/watch?v=Byv311hdoHE",
        "feed7_url": "https://feed7.dev/p/the-dirty-secret-of-forward-deployed-engineering-natalie-meurer-sierra-17crz97",
        "reason": "Both show agents compressing the path from identified work to production changes, while Uber demonstrates that organization-wide use requires centralized infrastructure and governance beyond individual ownership."
      },
      {
        "title": "NTT DATA Group cuts incident analysis to 30 minutes with Codex",
        "source_name": "OpenAI",
        "source_url": "https://openai.com/index/ntt-data",
        "feed7_url": "https://feed7.dev/p/ntt-data-1sgidqg",
        "reason": "NTT DATA provides a measured task-level time improvement, whereas Uber adds organization-scale PR and migration evidence; together they show different levels of enterprise adoption without proving product value."
      },
      {
        "title": "CFOs and the new economics of AI",
        "source_name": "Cursor",
        "source_url": "https://cursor.com/blog/cfo-council",
        "feed7_url": "https://feed7.dev/p/cfo-council-10ctxbn",
        "reason": "Cursor’s model-cost variance makes Uber’s standardized model and tool gateway an economic control point as well as an access and governance layer."
      }
    ]
  },
  "lifecycle": "Current",
  "published_at": "2026-08-21T13:00:06.000Z",
  "modified_at": "2026-08-21T13:00:06.000Z",
  "supersedes": [],
  "expires_at": null,
  "formats": {
    "html": "https://feed7.dev/p/agentic-sdlc-at-uber-uday-kiran-medisetty-adam-huda-uber-1ugtaxn",
    "json": "https://feed7.dev/p/agentic-sdlc-at-uber-uday-kiran-medisetty-adam-huda-uber-1ugtaxn.json",
    "markdown": "https://feed7.dev/p/agentic-sdlc-at-uber-uday-kiran-medisetty-adam-huda-uber-1ugtaxn.md"
  }
}