--- schema: 1 type: video title: 'Can LLMs generate Enterprise Quality Code? — Prasenjit Sarkar, Sonar' description: >- Sonar ran 4,444 Java programming assignments through 53 models and measured what actually came out. GPT-4o generated under 250,000 lines for those assignments. GPT 5.4 generated 1.2 million. Claude Sonnet 4.6 generated 627,000 with the highest security issue rate at 300 per million lines of code. Prasenjit Sarkar from Sonar walks through the full leaderboard: pass rate, cyclomatic complexity, bug density, and security issues per model. Their response is a three-stage framework called ACDC: guide, verify, solve. The verify stage runs SonarQube analysis in 1 to 5 seconds before a commit, against 1 to 5 minutes in CI. If issues slip through to the PR, a remediation agent creates one fix per issue, runs it through analysis and compilation to check for regressions, and only presents it if it passes. Speaker info: - https://www.linkedin.com/in/jit2600/ Timestamps: 0:00 Introduction and the Shift to Agentic Development 1:44 Evaluating LLM Code Quality and Reliability 3:00 Sonar's Evaluation Framework and Methodology 3:39 LLM Performance Analysis (Pass Rates and Code Bloat) 5:24 Why LLMs Struggle: Training Data and Hidden Flaws 6:45 The Sonar LLM Leaderboard 8:30 Complexity Metrics: Cyclomatic vs. Cognitive 10:41 The ACDC Framework: Guide, Verify, and Solve 11:06 Phase 1: Guide (Context Augmentation & Sonar Sweep) 11:42 Phase 2: Verify (SonarQube Agentic Analysis) 12:40 Phase 3: Solve (Remediation Agent) 14:05 Product Summary and Ecosystem Support --- Watch along with the corrected transcript, summary, timestamps and resources on this talk's official AIE page: https://ai.engineer/talks/NuePCNMpWGc-can-llms-generate-enterprise-quality-code-prasenjit-sarkar-sonar Join our upcoming AI Engineer conferences (see https://ai.engineer landing page): - AIEi Shanghai, November 5–6 — the first for meeting top Chinese labs and engineers - AIE CODE in SF, November 10–12 — the top AI Coding conference returns - AIEi Sydney, December 7–8 — the first AIEi run alongside NeurIPS 2026 in Sydney --- canonical: 'https://liquidgold.cc/watch/NuePCNMpWGc' markdown: 'https://liquidgold.cc/watch/NuePCNMpWGc.md' llms: 'https://liquidgold.cc/watch/NuePCNMpWGc/llms.txt' parent: 'https://liquidgold.cc/channel/aidotengineer.md' channel: aidotengineer updated: '2026-10-24T10:07:35.434Z' language: null children: [] videoId: NuePCNMpWGc durationSeconds: 912 publishedAt: '2026-05-31T18:00:21.000Z' searchChunks: - ordinal: 0 text: null startSeconds: 28 endSeconds: 105 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=28' heldReason: rights - ordinal: 1 text: null startSeconds: 28 endSeconds: 105 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=28' heldReason: rights - ordinal: 2 text: null startSeconds: 105 endSeconds: 116 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=105' heldReason: rights - ordinal: 3 text: null startSeconds: 116 endSeconds: 183 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=116' heldReason: rights - ordinal: 4 text: null startSeconds: 116 endSeconds: 183 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=116' heldReason: rights - ordinal: 5 text: null startSeconds: 183 endSeconds: 194 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=183' heldReason: rights - ordinal: 6 text: null startSeconds: 194 endSeconds: 226 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=194' heldReason: rights - ordinal: 7 text: null startSeconds: 226 endSeconds: 238 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=226' heldReason: rights - ordinal: 8 text: null startSeconds: 238 endSeconds: 328 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=238' heldReason: rights - ordinal: 9 text: null startSeconds: 238 endSeconds: 328 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=238' heldReason: rights - ordinal: 10 text: null startSeconds: 328 endSeconds: 340 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=328' heldReason: rights - ordinal: 11 text: null startSeconds: 340 endSeconds: 405 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=340' heldReason: rights - ordinal: 12 text: null startSeconds: 340 endSeconds: 405 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=340' heldReason: rights - ordinal: 13 text: null startSeconds: 405 endSeconds: 417 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=405' heldReason: rights - ordinal: 14 text: null startSeconds: 510 endSeconds: 521 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=510' heldReason: rights - ordinal: 15 text: null startSeconds: 521 endSeconds: 650 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=521' heldReason: rights - ordinal: 16 text: null startSeconds: 521 endSeconds: 650 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=521' heldReason: rights - ordinal: 17 text: null startSeconds: 650 endSeconds: 663 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=650' heldReason: rights - ordinal: 18 text: null startSeconds: 674 endSeconds: 685 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=674' heldReason: rights - ordinal: 19 text: null startSeconds: 685 endSeconds: 706 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=685' heldReason: rights - ordinal: 20 text: null startSeconds: 716 endSeconds: 769 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=716' heldReason: rights - ordinal: 21 text: null startSeconds: 716 endSeconds: 769 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=716' heldReason: rights - ordinal: 22 text: null startSeconds: 780 endSeconds: 846 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=780' heldReason: rights - ordinal: 23 text: null startSeconds: 780 endSeconds: 846 watch: 'https://liquidgold.cc/watch/NuePCNMpWGc?t=780' heldReason: rights --- # Can LLMs generate Enterprise Quality Code? — Prasenjit Sarkar, Sonar Sonar ran 4,444 Java programming assignments through 53 models and measured what actually came out. GPT-4o generated under 250,000 lines for those assignments. GPT 5.4 generated 1.2 million. Claude Sonnet 4.6 generated 627,000 with the highest security issue rate at 300 per million lines of code. Prasenjit Sarkar from Sonar walks through the full leaderboard: pass rate, cyclomatic complexity, bug density, and security issues per model. Their response is a three-stage framework called ACDC: guide, verify, solve. The verify stage runs SonarQube analysis in 1 to 5 seconds before a commit, against 1 to 5 minutes in CI. If issues slip through to the PR, a remediation agent creates one fix per issue, runs it through analysis and compilation to check for regressions, and only presents it if it passes. Speaker info: - https://www.linkedin.com/in/jit2600/ Timestamps: 0:00 Introduction and the Shift to Agentic Development 1:44 Evaluating LLM Code Quality and Reliability 3:00 Sonar's Evaluation Framework and Methodology 3:39 LLM Performance Analysis (Pass Rates and Code Bloat) 5:24 Why LLMs Struggle: Training Data and Hidden Flaws 6:45 The Sonar LLM Leaderboard 8:30 Complexity Metrics: Cyclomatic vs. Cognitive 10:41 The ACDC Framework: Guide, Verify, and Solve 11:06 Phase 1: Guide (Context Augmentation & Sonar Sweep) 11:42 Phase 2: Verify (SonarQube Agentic Analysis) 12:40 Phase 3: Solve (Remediation Agent) 14:05 Product Summary and Ecosystem Support --- Watch along with the corrected transcript, summary, timestamps and resources on this talk's official AIE page: https://ai.engineer/talks/NuePCNMpWGc-can-llms-generate-enterprise-quality-code-prasenjit-sarkar-sonar Join our upcoming AI Engineer conferences (see https://ai.engineer landing page): - AIEi Shanghai, November 5–6 — the first for meeting top Chinese labs and engineers - AIE CODE in SF, November 10–12 — the top AI Coding conference returns - AIEi Sydney, December 7–8 — the first AIEi run alongside NeurIPS 2026 in Sydney --- Canonical: https://liquidgold.cc/watch/NuePCNMpWGc · Channel: [AI Engineer](https://liquidgold.cc/channel/aidotengineer.md) > Search chunk text is held until public-source provenance is verified. ## Document chunks > Missing: no stored document chunks are available.