Agentic AI Reviews Cut Pull Request Time But Increase Review Smells · Digg

/AI

Agentic AI Reviews Cut Pull Request Time But Increase Review Smells

Combined views

5.6K

2 posts, first seen 29d ago

25 likes comments

Agentic AI Reviews Cut Pull Request Time But Increase Review Smells

Reactions

2 posts

RP
Rohan Paul@rohanpaul_ai 4 weeks ago
Agents can compress review time without improving the review process itself. A study of 1.02 million pull requests across 207 GitHub projects tracks the shift from human-only review to LLM-assisted and agentic review. Projects that adopted AI gradually, or moved rapidly to…
- likes: 25
- replies: 6
- bookmarks: 8
- reposts: 4
RP
Rohan Paul@rohanpaul_ai 4 weeks ago
Agents can compress review time without improving the review process itself. A study of 1.02 million pull requests across 207 GitHub projects tracks the shift from human-only review to LLM-assisted and agentic review. Projects that adopted AI gradually, or moved rapidly to…
- likes: 25
- replies: 6
- bookmarks: 8
- reposts: 4

Combined views

5.6K

2 posts, first seen 29d ago

25 likes

6

8 saves

6 reposts

6 comments

8 saves

6 reposts

Reactions

8 posts

RP
Rohan Paul@rohanpaul_ai 4 weeks ago
Agents can compress review time without improving the review process itself. A study of 1.02 million pull requests across 207 GitHub projects tracks the shift from human-only review to LLM-assisted and agentic review. Projects that adopted AI gradually, or moved rapidly to…
- likes: 25
- replies: 6
- bookmarks: 8
- reposts: 4
1. OW
  Owlfy.ai@Owlfy_ai 4 weeks ago
  @rohanpaul_ai Speeding things up can hide deeper process issues. Did they look at defect rates after merges, or just how fast reviews got done?
  - likes: 0
  - replies: 0
  - bookmarks: 0
  - reposts: 0
2. B+
  BowtiedWhitebat + Read Pinned Tweet or NGMI@bowtiedwhitebat 4 weeks ago
  @rohanpaul_ai it wilonly grow faster and be cheapaer
  - likes: 1
  - replies: 0
  - bookmarks: 0
  - reposts: 0
3. GR
  Gregor@bygregorr 4 weeks ago
  @rohanpaul_ai Not sure time compression and process quality are cleanly separable in the data. Did the gradual adopters show meaningfully different defect escape rates, or was most of the variance in throughput?
  - likes: 1
  - replies: 0
  - bookmarks: 0
  - reposts: 0
4. AN
  Anoy@Anoyroyc 4 weeks ago
  @rohanpaul_ai faster reviews with worse coverage is a bad trade
  - likes: 1
  - replies: 0
  - bookmarks: 0
  - reposts: 0
5. SA
  Saman Ahmed@ScaleWthAI 4 weeks ago
  @rohanpaul_ai The interesting challenge is not replacing reviewers, but designing better review systems around agents.
  - likes: 1
  - replies: 0
  - bookmarks: 0
  - reposts: 0