The AI Record
One of the biggest learnings in this process was like the model doesn't know that it's wrong. It always confidently says 100% sure that, yeah, this is the review for your code. Go ahead. But we saw that, no, it actually needs a lot of guidance from the teams, because each team has its own style guide, its own patterns or … anti-patterns that they want to look for.

Ameya KetkarSpeaker, AI Engineer World's Fair 2026 (job title not stated) · Practitioner, Uber

Context. Ameya Ketkar is describing what Uber's team learned while adding observability and evaluation to its uReview code-review agent.

Talk: Building uReview, Uber's Multi-Agent Code Review Engine (talk, AI Engineer World's Fair 2026), AI Engineer (YouTube), August 28, 2026, at 06:35.

Not yet checked [ Grade B ] Said in the last year Uber: $52 billion revenue, FY2025

Filed under

Who is speaking

Ameya Ketkar, Speaker, AI Engineer World's Fair 2026 (job title not stated). Describes the evaluation and rollout of a code-review system his team built and runs, so he is close to the numbers; Uber is promoting its own work.

How this was checked

Not yet checked by an independent second reader. The wording, speaker and date come from the source linked above; treat it as a lead until it is confirmed.

First reader (reader-youtube lane, 2026-11-01): both whisper passes (full-audio transcription and a separate re-transcription of the cut clip with 5 seconds of margin) agree on every word quoted. The second pass dropped the word "like" before "anti-patterns", so that word is cut with …. The second pass spells "100 percent"; "100%" is the normalization. Not yet checked by a blind second reader.

How to cite this passage

Ameya Ketkar (Uber), “Building uReview, Uber's Multi-Agent Code Review Engine (talk, AI Engineer World's Fair 2026)”, AI Engineer (YouTube), 2026-08-28, at 06:35. https://www.youtube.com/watch?v=EL123UNokkI Archived at https://the-ai-record.com/passages/uber-one-biggest-learnings-process-like-model-2026

More from the same source

Back in 2024, we were seeing that engineers would get their first review within three hours. Now, in 2026, that has grown to nine hours, in addition to all of the volume changes. So in short, code review is now the bottleneck that we are running into.

Will BondStaff Engineer, AI Foundations team, Development Platform organization · Practitioner, Uber

Context. Will Bond is describing a metric Uber tracks, the time to a first review of a pull request.

Talk: Building uReview, Uber's Multi-Agent Code Review Engine (talk, AI Engineer World's Fair 2026), AI Engineer (YouTube), August 28, 2026, at 01:05.

  • 3 hours first review time 2024, 2024
  • 9 hours first review time 2026, 2026

Not yet checked [ Grade B ] Said in the last year Uber: $52 billion revenue, FY2025 Passage page

Currently, we still have humans approving the code, but we see a short path in the near future to a percentage of our code landing automatically, having automatic approvals, right?

Will BondStaff Engineer, AI Foundations team, Development Platform organization · Practitioner, Uber

Context. Will Bond is describing where Uber expects code review to go as it moves to an agentic software development life cycle.

Talk: Building uReview, Uber's Multi-Agent Code Review Engine (talk, AI Engineer World's Fair 2026), AI Engineer (YouTube), August 28, 2026, at 11:39.

Not yet checked [ Grade B ] Said in the last year Uber: $52 billion revenue, FY2025 Passage page

One thing that we learned is that actually writing the skill was very easy, like teams just very quickly wrote a skill by asking Claude to write one, go over my previous PR reviews and write me a skill, but the hard part was how to run these skills at scale with consistent quality and low cost, and that required a lot of iterations, not only from the [uReview] team side, but also … for each team who was trying to write these rules.

Ameya KetkarSpeaker, AI Engineer World's Fair 2026 (job title not stated) · Practitioner, Uber

Context. Ameya Ketkar is describing how Uber's teams wrote their own code-review skills for uReview.

Talk: Building uReview, Uber's Multi-Agent Code Review Engine (talk, AI Engineer World's Fair 2026), AI Engineer (YouTube), August 28, 2026, at 09:44.

Not yet checked [ Grade B ] Said in the last year Uber: $52 billion revenue, FY2025 Passage page

But rather than killing the outer loop, I think that we believe and the industry has just started to really kind of coalesce on this idea that … really expanding the outer loop. Rather than removing humans from the code review process, we are moving their responsibilities up a layer.

Will BondStaff Engineer, AI Foundations team, Development Platform organization · Practitioner, Uber

Context. Will Bond is closing the talk on what happens to human reviewers as agents write and review more of the code.

Talk: Building uReview, Uber's Multi-Agent Code Review Engine (talk, AI Engineer World's Fair 2026), AI Engineer (YouTube), August 28, 2026, at 13:43.

Not yet checked [ Grade B ] Said in the last year Uber: $52 billion revenue, FY2025 Passage page

Uber spent the time to invest in building an in-house solution due to some of the constraints that we have. One of those is we currently use [Phabricator] and have for a long time and are in the process of migrating to GitHub. Most of the solutions do not provide support for [Phabricator].

Will BondStaff Engineer, AI Foundations team, Development Platform organization · Practitioner, Uber

Talk: Building uReview, Uber's Multi-Agent Code Review Engine (talk, AI Engineer World's Fair 2026), AI Engineer (YouTube), August 28, 2026, at 01:22.

Not yet checked [ Grade B ] Said in the last year Uber: $52 billion revenue, FY2025 Passage page

[uReview] does … around 25,000 comments a week, and we get 10% of them actually get some feedback … We also saw that the overall address rate was around 67%, and almost three quarters of the high severity issues were usually addressed by the developers

Ameya KetkarSpeaker, AI Engineer World's Fair 2026 (job title not stated) · Practitioner, Uber

Context. Ameya Ketkar is reporting results for uReview, Uber's code-review system. The address rate is how often a developer acts on a comment, as he defined it earlier in the talk.

Talk: Building uReview, Uber's Multi-Agent Code Review Engine (talk, AI Engineer World's Fair 2026), AI Engineer (YouTube), August 28, 2026, at 10:13.

  • 25,000 comments ai review comments, per week. AI assists a person
  • 67% review comment address rate, as of the talk. AI assists a person

Not yet checked [ Grade B ] Said in the last year Uber: $52 billion revenue, FY2025 Passage page

we saw that against like a very naive implementation, our costs were down by 60%, and our quality and our accuracy was up by around 70%.

Ameya KetkarSpeaker, AI Engineer World's Fair 2026 (job title not stated) · Practitioner, Uber

Context. Ameya Ketkar is summarizing the effect of the observability and evaluation work on uReview, Uber's code-review system.

Talk: Building uReview, Uber's Multi-Agent Code Review Engine (talk, AI Engineer World's Fair 2026), AI Engineer (YouTube), August 28, 2026, at 10:50.

  • -60% ai review cost change, against a naive first implementation
  • 70% ai review quality change, against a naive first implementation

Not yet checked [ Grade B ] Said in the last year Uber: $52 billion revenue, FY2025 Passage page

Something wrong? Report a correction. JSON: /api/v1/passages/uber-one-biggest-learnings-process-like-model-2026.json