← All implementations

Uber · Background agent

uReview

An event-driven AI code reviewer for Uber's internal review platform that generates, grades, filters, deduplicates, and posts findings while leaving engineers in control of the reviewed change.

1

Approach type
Background agent
Work
Code review
Human involvement
Drafts reviewed
Invocation
Background, Event driven
Interfaces
Internal ui, Ci
Deployment stage
Scaled
Evidence strength
Detailed primary
Entry reviewed

How it works

The workflow the sources report for this implementation.

Standard, best-practices, and AppSec reviewers generate findings for different issue classes

1

Confidence grading, semantic deduplication, and historically low-value category suppression reduce noise

1

Developer ratings, addressed-comment detection, and a curated benchmark tune prompts, thresholds, and models

1

Where people stay involved

  • pull request → filtered AI review findingsWork product review · Level 3

Level 3 for pull request → filtered AI review findings; human attention boundary: work-product-review.

1

Observation date
2025-08-12

Implementation details

Harness

A prompt-chained pipeline separates comment generation, confidence grading, validation, semantic deduplication, and category filtering; three specialized assistants were in operation when published

1

Model

Periodic benchmark evaluation; Claude 4 Sonnet as generator with o4-mini-high as grader was the highest-F1 reported pairing

1

Interfaces

internal-ui, ci

1

Tool access

Reviews eligible code in Uber's six monorepos across Go, Java, Android, iOS, TypeScript, and Python; richer internal artifacts were not yet connected when published

1

Knowledge

Surrounding source context plus a shared registry of Uber-specific coding and style rules

1

Context management

Comments and metadata are streamed through Kafka to Hive for feedback analysis, experiments, and operational dashboards

1

Reported results and limitations

The catalog records what the sources report, with the scope and the denominator of every figure. A qualification below limits the figure it sits under.

Reported metrics

Headline claim

Uber's introduction reports reviews of over 90% of approximately 65,000 weekly diffs, with over 75% usefulness and over 65% addressed comments; a later paragraph says 65,000 diffs per month

1 Supports1 Contradicts

Conflicting evidence

The preserved evidence disagrees about this statement. Read the qualifications above and source 1.

Reported by
Uber
Scope
Introduction's reported weekly diff coverage, plus comment usefulness/addressed rates; later paragraph's monthly period conflicts and remains unresolved
Denominator
Introduction reports approximately 65,000 weekly diffs; cost paragraph says monthly. Usefulness covers rated comments; addressed rate covers posted comments
Method
Production coverage, developer ratings, and automatic addressed-comment detection
Observation date
2025-08
Key observation

Introduction reports reviews of over 90% of approximately 65,000 weekly diffs; cost discussion instead says 65,000 diffs per month

1 Supports1 Contradicts

Conflicting evidence

The preserved evidence disagrees about this statement. Read the qualifications above and source 1.

Reported by
Uber
Scope
Introduction's reported weekly diffs analyzed; monthly period in the cost paragraph conflicts and remains unresolved
Denominator
Approximately 65,000 weekly diffs according to the introduction; same volume described as monthly in the cost paragraph
Observation date
2025-08
Key observation

Over 75% of comments rated useful by engineers who interact with the tool

1

Reported by
Uber
Scope
Comments rated useful by engineers who provide feedback
Denominator
Comments with engineer interaction
Method
Useful / Not Useful rating links
Observation date
2025-08
Key observation

Over 65% of posted comments addressed in the same changeset

1

Reported by
Uber
Scope
Posted comments considered addressed in the same changeset
Denominator
Posted comments
Method
Five reruns on the final commit and semantic-similarity matching
Observation date
2025-08
Key observation

Median review latency of 4 minutes across all six Uber monorepos

1

The source does not report the denominator of this figure.

Reported by
Uber
Scope
Reviews across all six Uber monorepos
Method
Production latency telemetry
Observation date
2025-08
Key observation

Approximately 1,500 developer hours reportedly saved per week, based on an assumed 10-minute second review per processed commit

1

Reported by
Uber
Scope
Processed commits excluding configuration files; modeled second-review time savings
Denominator
Over 10,000 commits per week
Method
Processed commits multiplied by an assumed 10 minutes for a second human review
Observation date
2025-08

Lessons and interpretation

Prefer fewer high-confidence findings over high comment volume

1

Combine prompts with deterministic filtering, deduplication, evaluation, and feedback instrumentation

1

Roll out gradually by team and assistant while tracking precision, recall, usefulness, and false positives

1

Sources and research details

Citations link to the original publisher. Each source also keeps a preserved copy in the repository, so a changed or removed page stays checkable.

  1. uReview: Scalable, Trustworthy GenAI for Code Review at Uberhttps://www.uber.com/us/en/blog/ureview/Engineering blog · First party · Last source verification: 2026-08-31
    Preserved copy ↗
Research details for every claim on this page
  1. Summary
    Statement type
    Fact
    Provenance
    Reported
    Confidence
    High
    Confidence reason
    A linked first-party source states the claim.
  2. Headline claim
    Statement type
    Metric
    Provenance
    Reported
    Confidence
    Low
    Confidence reason
    The opening paragraph reports approximately 65,000 weekly diffs, but the cost discussion says 65,000 per month. These conflicting periods remain unresolved; neither is independently verified.
    Reported by
    Uber
    Scope
    Introduction's reported weekly diff coverage, plus comment usefulness/addressed rates; later paragraph's monthly period conflicts and remains unresolved
    Denominator
    Introduction reports approximately 65,000 weekly diffs; cost paragraph says monthly. Usefulness covers rated comments; addressed rate covers posted comments
    Method
    Production coverage, developer ratings, and automatic addressed-comment detection
    Observation date
    2025-08
  3. Harness
    Statement type
    Fact
    Provenance
    Reported
    Confidence
    High
    Confidence reason
    A linked first-party source states the claim.
  4. Model
    Statement type
    Fact
    Provenance
    Reported
    Confidence
    High
    Confidence reason
    A linked first-party source states the claim.
  5. Interfaces
    Statement type
    Fact
    Provenance
    Reported
    Confidence
    High
    Confidence reason
    A linked first-party source states the claim.
  6. Tool access
    Statement type
    Fact
    Provenance
    Reported
    Confidence
    High
    Confidence reason
    A linked first-party source states the claim.
  7. Knowledge
    Statement type
    Fact
    Provenance
    Reported
    Confidence
    High
    Confidence reason
    A linked first-party source states the claim.
  8. Context management
    Statement type
    Fact
    Provenance
    Reported
    Confidence
    High
    Confidence reason
    A linked first-party source states the claim.
  9. Supporting component
    Statement type
    Fact
    Provenance
    Reported
    Confidence
    High
    Confidence reason
    A linked first-party source states the claim.
  10. Supporting component
    Statement type
    Fact
    Provenance
    Reported
    Confidence
    High
    Confidence reason
    A linked first-party source states the claim.
  11. Supporting component
    Statement type
    Fact
    Provenance
    Reported
    Confidence
    High
    Confidence reason
    A linked first-party source states the claim.
  12. Key observation
    Statement type
    Metric
    Provenance
    Reported
    Confidence
    Low
    Confidence reason
    The opening paragraph reports approximately 65,000 weekly diffs, but the cost discussion says 65,000 per month. These conflicting periods remain unresolved; neither is independently verified.
    Reported by
    Uber
    Scope
    Introduction's reported weekly diffs analyzed; monthly period in the cost paragraph conflicts and remains unresolved
    Denominator
    Approximately 65,000 weekly diffs according to the introduction; same volume described as monthly in the cost paragraph
    Method
    Not reported
    Observation date
    2025-08
  13. Key observation
    Statement type
    Metric
    Provenance
    Reported
    Confidence
    Medium
    Confidence reason
    A linked first-party source states the claim.
    Reported by
    Uber
    Scope
    Comments rated useful by engineers who provide feedback
    Denominator
    Comments with engineer interaction
    Method
    Useful / Not Useful rating links
    Observation date
    2025-08
  14. Key observation
    Statement type
    Metric
    Provenance
    Reported
    Confidence
    Medium
    Confidence reason
    A linked first-party source states the claim.
    Reported by
    Uber
    Scope
    Posted comments considered addressed in the same changeset
    Denominator
    Posted comments
    Method
    Five reruns on the final commit and semantic-similarity matching
    Observation date
    2025-08
  15. Key observation
    Statement type
    Metric
    Provenance
    Reported
    Confidence
    Medium
    Confidence reason
    A linked first-party source states the claim.
    Reported by
    Uber
    Scope
    Reviews across all six Uber monorepos
    Denominator
    Not reported
    Method
    Production latency telemetry
    Observation date
    2025-08
  16. Key observation
    Statement type
    Metric
    Provenance
    Reported
    Confidence
    Low
    Confidence reason
    The source estimates about 1,500 hours using over 10,000 commits and a 10-minute assumption; the rounded figures do not arithmetically reconcile exactly and are not observed time savings.
    Reported by
    Uber
    Scope
    Processed commits excluding configuration files; modeled second-review time savings
    Denominator
    Over 10,000 commits per week
    Method
    Processed commits multiplied by an assumed 10 minutes for a second human review
    Observation date
    2025-08
  17. Lesson
    Statement type
    Inference
    Provenance
    Catalog judgment
    Confidence
    High
    Confidence reason
    The article states this lesson directly and documents the associated mechanisms.
  18. Lesson
    Statement type
    Inference
    Provenance
    Catalog judgment
    Confidence
    High
    Confidence reason
    The article states this lesson directly and documents the pipeline.
  19. Lesson
    Statement type
    Inference
    Provenance
    Catalog judgment
    Confidence
    High
    Confidence reason
    The article states this lesson directly and describes the rollout telemetry.
  20. Operating model assessment
    Statement type
    Inference
    Provenance
    Catalog judgment
    Confidence
    High
    Confidence reason
    The source describes AI-generated review findings while engineers remain in control of the change.
    Observation date
    2025-08-12
Back to the directory →