--- schema: 1 type: video title: 'From coding to Knowledge work agents — Karan Vaidya, Composio' description: >- Karan Vaidya pointed his own OpenClaw at hiring outreach and it mass emailed candidates exactly as instructed. Some of the people in the room had received one. The thread that followed put his name on Twitter, and every check in the software engineering playbook would have passed. The addresses were real, the emails were valid, they reached actual people. Nothing tested the only question that mattered, which was whether the outreach should have gone at all. Vaidya, cofounder and CTO of Composio, turns his own disaster into a larger claim: coding agents did not race ahead because models are better at code, but because code already had everything an agent needs around it. He walks six primitives coding got for free and knowledge work lacks entirely. Centralization, since one deal is scattered across Salesforce, Notion, Gmail, Slack and Zendesk behind five separate logins. History, because git records every change while business apps keep nothing an agent can read back. Context, both the map of how an organization works and the unwritten sense of what good looks like. Then verification, governance and reversibility. On governance he cites the alignment director at Meta whose email agent kept deleting messages after she told it to stop, until she reached a physical machine. Two hundred were gone, and she had prompted it in advance to confirm. A prompt lives in memory that gets compacted away, so Composio puts the wall outside the agent. Reversibility is the one he concedes is hardest, because a sent email or a wire cannot be called back. Speaker info: - https://x.com/KaranVaidya6 - https://www.linkedin.com/in/kaavee315/ - https://kvaidya.com/ Timestamps: 0:00 - Why nearly all agent tool calls are software engineering 0:43 - From autocomplete to autonomous in three years 1:26 - The scaffolding code already had 2:35 - Centralization, and one deal across five apps 4:11 - History, and agents that start blank every time 6:55 - Two kinds of context, the map and the style 9:26 - Verification, and how code checks its own work 10:25 - The outreach agent that ended up on Twitter 12:31 - Governance, and 200 emails deleted anyway 15:01 - Walls the agent cannot argue with 16:24 - Reversibility, and work that has no undo 19:09 - The bottleneck moves off the model --- Watch along with the corrected transcript, summary, timestamps and resources on this talk's official AIE page: https://ai.engineer/talks/xxfMT-bPEmU-youtube-daf213ba1da9e42198d6 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/xxfMT-bPEmU' markdown: 'https://liquidgold.cc/watch/xxfMT-bPEmU.md' llms: 'https://liquidgold.cc/watch/xxfMT-bPEmU/llms.txt' parent: 'https://liquidgold.cc/channel/aidotengineer.md' channel: aidotengineer updated: '2026-10-24T04:00:06.589Z' language: null children: [] videoId: xxfMT-bPEmU durationSeconds: 1242 publishedAt: '2026-09-03T16:00:06.000Z' searchChunks: - ordinal: 0 text: null startSeconds: 13 endSeconds: 24 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=13' heldReason: rights - ordinal: 1 text: null startSeconds: 45 endSeconds: 60 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=45' heldReason: rights - ordinal: 2 text: null startSeconds: 60 endSeconds: 93 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=60' heldReason: rights - ordinal: 3 text: null startSeconds: 93 endSeconds: 104 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=93' heldReason: rights - ordinal: 4 text: null startSeconds: 104 endSeconds: 161 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=104' heldReason: rights - ordinal: 5 text: null startSeconds: 173 endSeconds: 251 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=173' heldReason: rights - ordinal: 6 text: null startSeconds: 263 endSeconds: 421 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=263' heldReason: rights - ordinal: 7 text: null startSeconds: 263 endSeconds: 421 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=263' heldReason: rights - ordinal: 8 text: null startSeconds: 421 endSeconds: 432 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=421' heldReason: rights - ordinal: 9 text: null startSeconds: 432 endSeconds: 566 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=432' heldReason: rights - ordinal: 10 text: null startSeconds: 577 endSeconds: 629 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=577' heldReason: rights - ordinal: 11 text: null startSeconds: 640 endSeconds: 751 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=640' heldReason: rights - ordinal: 12 text: null startSeconds: 640 endSeconds: 751 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=640' heldReason: rights - ordinal: 13 text: null startSeconds: 762 endSeconds: 912 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=762' heldReason: rights - ordinal: 14 text: null startSeconds: 912 endSeconds: 922 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=912' heldReason: rights - ordinal: 15 text: null startSeconds: 922 endSeconds: 992 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=922' heldReason: rights - ordinal: 16 text: null startSeconds: 992 endSeconds: 1005 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=992' heldReason: rights - ordinal: 17 text: null startSeconds: 1005 endSeconds: 1151 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=1005' heldReason: rights - ordinal: 18 text: null startSeconds: 1151 endSeconds: 1163 watch: 'https://liquidgold.cc/watch/xxfMT-bPEmU?t=1151' heldReason: rights --- # From coding to Knowledge work agents — Karan Vaidya, Composio Karan Vaidya pointed his own OpenClaw at hiring outreach and it mass emailed candidates exactly as instructed. Some of the people in the room had received one. The thread that followed put his name on Twitter, and every check in the software engineering playbook would have passed. The addresses were real, the emails were valid, they reached actual people. Nothing tested the only question that mattered, which was whether the outreach should have gone at all. Vaidya, cofounder and CTO of Composio, turns his own disaster into a larger claim: coding agents did not race ahead because models are better at code, but because code already had everything an agent needs around it. He walks six primitives coding got for free and knowledge work lacks entirely. Centralization, since one deal is scattered across Salesforce, Notion, Gmail, Slack and Zendesk behind five separate logins. History, because git records every change while business apps keep nothing an agent can read back. Context, both the map of how an organization works and the unwritten sense of what good looks like. Then verification, governance and reversibility. On governance he cites the alignment director at Meta whose email agent kept deleting messages after she told it to stop, until she reached a physical machine. Two hundred were gone, and she had prompted it in advance to confirm. A prompt lives in memory that gets compacted away, so Composio puts the wall outside the agent. Reversibility is the one he concedes is hardest, because a sent email or a wire cannot be called back. Speaker info: - https://x.com/KaranVaidya6 - https://www.linkedin.com/in/kaavee315/ - https://kvaidya.com/ Timestamps: 0:00 - Why nearly all agent tool calls are software engineering 0:43 - From autocomplete to autonomous in three years 1:26 - The scaffolding code already had 2:35 - Centralization, and one deal across five apps 4:11 - History, and agents that start blank every time 6:55 - Two kinds of context, the map and the style 9:26 - Verification, and how code checks its own work 10:25 - The outreach agent that ended up on Twitter 12:31 - Governance, and 200 emails deleted anyway 15:01 - Walls the agent cannot argue with 16:24 - Reversibility, and work that has no undo 19:09 - The bottleneck moves off the model --- Watch along with the corrected transcript, summary, timestamps and resources on this talk's official AIE page: https://ai.engineer/talks/xxfMT-bPEmU-youtube-daf213ba1da9e42198d6 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/xxfMT-bPEmU · 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.