Advertise on podcast: Chain of Thought | AI Agents, Infrastructure & Engineering

What breaks when agents become the customer? Most of the integration layer, APIs, SDKs and webhooks alike, was built for human developers and for servers that are always listening. Agents are neither. Full transcript and chapters: https://chainofthought.show/podcast/78-what-breaks-when-agents-become-the-customer-svix-ceo-tom-hacohen/ When an agent runs on a laptop, closing the lid takes down the tunnel its webhooks come through. When the lid opens, it gets a storm of retries from 12 hours earlier. Svix founder and CEO Tom Hacohen's answer is to let the agent poll for what it missed. Svix delivers webhooks for companies like Clerk and Resend, and it created Standard Webhooks, the spec OpenAI, Anthropic and Google Gemini have adopted. In this episode of Chain of Thought, Tom joins Conor Bronsdon to talk through designing APIs for agents that adapt, whether agents still need SDKs, and why he wants less code even when agents write it. He also introduces Poke, Svix's new spec for making event-driven APIs work for agents. This episode is sponsored by Svix. We cover: Why events are harder for agents than API calls: the sender controls them, and they arrive outside the agent's sessionHow polling endpoints and Svix's plugins for Hermes and OpenClaw let an agent pick up where it left offWhy Tom calls JSON an abomination and webhooks duct tape, and why worse is better anywayWhy Svix created Standard Webhooks when everyone implemented webhooks differently, and how the spec let Kong and ngrok add webhook verificationWhether agents still need SDKs, and why Svix's skills work like a solutions engineer while its docs stay written for humansWhy Tom wants less code even when agents write it, and what piling patches on patches costsPoke, Svix's early spec for letting agents subscribe to and trigger events across services, released so feedback can shape it while it's still malleableChapters:(0:00) Duct tape, retry storms and the case for simplicity(1:19) Why integrations got easier and harder at once(2:59) How EteSync's webhook requests led to Svix(5:22) Worse is better: why webhooks persist(7:15) What changes when an agent receives the event(9:07) What happens when the laptop closes(11:08) Hermes, OpenClaw and agents that poll(13:03) Letting agents choose their own subscriptions(14:42) The long tail of the web without APIs(18:38) Why Svix created Standard Webhooks(24:17) How Standard Webhooks is governed(26:16) Following real usage, including MCP(28:17) Designing APIs for agents that adapt(30:38) Do agents still need SDKs?(31:25) Skills as documentation for agents(34:52) Less code, even when agents write it(36:19) Agentic development and the offshoring analogy(37:34) Software factories and the plumbing underneath(40:57) Keeping up while the standards change(44:18) What Svix is launching with Poke(46:39) Why Svix released the Poke spec early(47:58) Building for your customers' customers(49:38) Clerk, Resend and webhooks in the background(52:22) Where to follow Tom Links from the episode: Svix: https://link.svix.com/cotPoke: https://www.svix.com/poke/Standard Webhooks: https://www.standardwebhooks.comStandard Webhooks on GitHub: https://github.com/standard-webhooks/standard-webhooksEteSync: https://www.etesync.comssot-check: https://github.com/conorbronsdon/ssot-checkMost of the Web Will Never Get APIs for AI Agents | Dhruv Batra: https://chainofthought.show/podcast/63-most-of-the-web-will-never-get-apis-for-ai-agents-dhruv-batra/Connect with Tom Hacohen: LinkedIn: https://www.linkedin.com/in/tomhacohen/X: https://x.com/TomHacohenSvix: https://link.svix.com/cotConnect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdonMore episodes: https://chainofthought.show Thanks to Svix, presenting sponsor of season four of Chain of Thought. Svix delivers billions of reliable webhooks for startups and the Fortune 500. Get started at https://link.svix.com/cot. Qualified startups will get $12,000 in credits, $50,000 for YC companies.

If your agent's memory lives inside one model provider, switching models or harnesses means starting over. You need portable memory - yet when switching agent memory between models, you can swing accuracy by thirteen points or more, depending on which model reads them back. Full transcript and chapters: https://chainofthought.show/podcast/77-switching-models-shouldnt-mean-starting-over-walrus-protocols-kimberly-logan/ Kimberly Logan, Head of Product at the Walrus Foundation, where she builds Walrus Memory, joins Chain of Thought to discuss how to build portable and verifiable memory for AI agents - and why Walrus's decentralized storage network has proved an apt primitive. She and host Conor Bronsdon get into why model lock-in is a memory problem, what people use agent memory for beyond coding, how you prove what happened to your data, and why Walrus doesn't publish memory quality benchmarks, plus more. This episode is sponsored by Walrus. We cover: Why longer context windows don't help once you switch models or harnesses, and why an import function only fixes one point in timeWhy Kimberly doesn't benchmark Walrus Memory against memory startups, and why she sees the opportunity as bigger than memory aloneWhat people build with Walrus Memory besides coding agents, which Walrus puts at maybe 30% of usage: long-running study aids, trading agents, and security teams comparing production snapshots with code changesWalrus's figure that almost 85% of the memories written are read back more than 30 days laterHow attestations on the blockchain and a delegate key you control let you show what happened to your data, and why unchanged data still isn't proof that it's accurateWhy Kimberly treats recall quality and latency as table stakes, and how a knowledge graph keeps recall on the freshest memoriesWhy she left almost 20 years in traditional software, including seven at Google, for a blockchain company, and where shared memory across companies and trust boundaries goes nextWalrus Console, announced this week: one place to manage agent memories and files through an interface or an MCP connection, without knowing how the storage underneath is configuredConnect with Kimberly Logan: LinkedIn: https://www.linkedin.com/in/krlgn/Twitter/X: https://x.com/kimblgnWalrus: https://walrus.xyz/cotConnect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdon🔗 More episodes: https://chainofthought.show Chapters:(0:00) Why agent memory has to move with you(1:25) Context windows don't follow you to a new model(4:41) Why an import function doesn't fix lock-in(6:38) Why Walrus isn't competing head-on with memory startups(8:39) Where Walrus Memory goes next(10:38) Trading agents and security snapshots on Walrus Memory(14:44) Verifiability: answering what happened to your data(17:21) Fresh recall, and why Walrus doesn't publish memory benchmarks(20:22) From traditional software to a blockchain company(24:40) A record other companies and regulators can check(25:42) What Walrus Console is for(29:09) Use agent memory beyond coding agents(30:52) Plugging in your own agent stack(31:41) Memory across companies and trust boundaries(35:48) Closing thoughts Thanks to Walrus, presenting sponsor of season four of Chain of Thought. Walrus Memory gives AI agents portable, verifiable memory that carries context across apps, sessions, and other agents. Get started at https://walrus.xyz/cot #AI #AIAgents #AgentMemory #SoftwareEngineering #DevTools

Andrea Malagodi is CTO at Sonar, which builds AI code verification and governance tools. When finance asked about the size of his team's AI bills, he calculated cost per PR. At one extreme was an engineer with more than 500 PRs in a short period and a low cost per PR. At the other was an engineer with very high spend and very few PRs. Full transcript and chapters: https://chainofthought.show/podcast/76-your-best-ai-engineer-might-have-the-worst-metrics-sonar-cto-andrea-malagodi/ Judged on cost alone, the first engineer wins and the second looks like a problem. Malagodi looked closer. The first had built a personal agent factory, with specification, design review, documentation records, coding, verification and functional testing, and about two-thirds of the output was tests and validations. The second was working on a hard problem that needed the AI to reason through long, multi-turn sessions and didn't reduce to a line count. Numbers alone tell you something, but not which engineer was doing the more valuable work. We cover: Why cost per PR is becoming the industry's standard AI spend metric, and how Malagodi's 500-PR engineer and high-spend engineer broke itThe history of bad proxies, from lines of code to story points to token leaderboards, and why Malagodi says the people who removed code were often the most importantSonar's guide, verify, solve loop: Vortex for codebase context, algorithmic plus LLM-based PR analysis, a remediation agent for legacy debt, and the Hunter agent for security vulnerabilitiesWhy Malagodi breaks agent work into small pieces, and why a five-day session that produces 400,000 lines of code leaves you with no idea what is in your codebaseSetting up disagreement between agents with an orchestrator and personas for engineering, product management and quality, plus Conor Bronsdon's cross-model-family review setupRolling AI tools out at enterprise scale: one cost view for finance, up to 30% savings from cutting repeated context reads, and why budgets as small as $50 a week are hard to work withRotating a triage duty to protect deep work, including how Sonar's teams built skills to sort through hundreds of reported CVEs, a number of them likely false positivesNew episodes, the ideas behind them and Conor's essays land in the Chain of Thought newsletter first. Subscribe: https://newsletter.chainofthought.show/ Chapters:(0:00) Human reviewers rubber-stamp big AI changes(2:43) Learning the craft alongside AI tools(9:09) From lines of code to PR counts(12:17) Spreading top engineers' setups org-wide(16:50) The guide, verify, solve agent loop(22:51) Specialized agent lanes and multitasking limits(26:51) Clear asks and guardrails for agent coordination(30:29) Building disagreement between agents(34:59) Mixing models instead of picking one(39:46) Rolling out AI tooling at enterprise scale(46:19) Defense in depth for AI-written code(50:52) Protecting time for incident investigations(54:08) Personalized software for teams and individuals Links from the episode: Sonar Vortex research: https://www.sonarsource.com/blog/cut-your-coding-agents-cost-with-sonar-semantic-code-navigation/SonarQube Hunter Agent: https://www.sonarsource.com/products/sonarqube/hunter-agent/Connect with Andrea Malagodi: LinkedIn: https://www.linkedin.com/in/malagodia/Sonar: https://www.sonar.com/Connect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdonMore episodes: https://chainofthought.show Thanks to Svix, presenting sponsor of season four of Chain of Thought. Svix delivers billions of reliable webhooks for startups and the Fortune 500. Get started at https://link.svix.com/cot. Qualified startups get $12,000 in credits, and YC companies get $50,000. Thanks to Walrus, presenting sponsor of season four of Chain of Thought. Walrus Memory gives AI agents portable, verifiable memory that carries context across apps, sessions, and other agents. Get started at https://walrus.xyz/cot Thanks to Inngest, presenting sponsor of season four of Chain of Thought. Agents in production run long - they call models and wait on APIs and people. But the longer agents run, the more they break. Inngest handles that with durable execution. You build your agent as steps in TypeScript, Python, or Go. When a step fails, Inngest retries it with exponential backoff, and completed steps are saved and skipped. Try it out: https://inngest.link/cot-pod Thanks to G2i for sponsoring this episode - for over a decade, they vetted and placed engineers at other companies, from startups to FAANG. Two years ago, they turned that same judgment inward, building their own bench to review RL environments, evals, and training data that models are trained on. Get access: https://fandf.co/3SFxVm6

An AI agent answering questions about your business needs to understand how that business works. How do you calculate churn? When does your fiscal year end? Who has the authority to change those definitions? Connecting a model to company data leaves those questions unresolved. Full transcript and chapters: https://chainofthought.show/podcast/75-why-context-alone-isnt-enough-for-enterprise-ai-agents-wisdomai-cpo-kapil-chhabra/ That’s the work of context engineering: giving agents the business definitions, instructions and examples they need to interpret your data. But that context changes as the business evolves, and someone has to keep it accurate. In this episode of Chain of Thought, WisdomAI co-founder and CPO Kapil Chhabra joins Conor Bronsdon to explain how his team approaches that challenge. We explore how companies maintain shared context, why data teams are taking on the role of AI context engineers, and how a specialized harness plans queries, checks results and repairs errors before returning an answer. Recorded at WisdomAI’s San Mateo office, this episode is sponsored by WisdomAI. We cover: Why the context layer is broader than a semantic layer or catalog, and how it differs from memoryHow context drifts, and why a subject-matter expert has to approve changes to shared definitionsWhy Kapil sees an AI context engineer role emerging, and why data teams are moving from providing insights to providing contextHow WisdomAI's harness decomposes a question, federates queries across data sources and repairs errors, including a customer that replaced a $5M-a-year analytics pipelineThe four ingredients of a trustworthy AI answer: accuracy, consistency, governance and explainabilityLive Apps: governed analytics apps built from a single prompt, and what keeps them liveWhy your context is your IP and should stay portableChapters:(0:00) Do your agents have the right context?(1:56) The four ingredients of trust(5:13) The criticality and impact 2x2(8:12) Data, context, harness: the hospital analogy(11:12) What the context layer actually means(11:48) Specialized harnesses: legal, support, analytics(13:17) Why only 7% of data leaders have scaled AI(15:46) What models can't guess: ARR, churn, fiscal years(16:45) The data stack collapses into the context layer(20:55) Memory vs. context(25:26) Are agents the new users of software?(27:19) Where humans should spend their time(28:14) Commissioning an AI agent, and who verifies it(31:28) Context drift and the learning loop(34:04) Context is a multiplayer game(35:12) Decompose, query, verify, repair(38:47) Replacing a $5M analytics pipeline with federation(42:04) The context development life cycle(43:29) The AI context engineer(46:02) Jobs are changing, not disappearing(46:59) Product, people and process(51:20) Who decides? Why FDEs can't own your context(52:22) Data context vs. business context(54:11) The benchmark: specialized harness vs. general agent(56:08) Meeting users in ChatGPT, Claude and Slack(58:48) Static vs. runtime context(1:00:08) Harness engineering as models change(1:01:44) Right-sizing AI and Live Apps(1:04:42) The boring parts: governance, security, caching(1:06:18) 1,000 dashboards, 50 human-years(1:08:08) What "live" means(1:09:19) Are dashboards going away?(1:12:08) A pipeline app built on a weekend walk(1:15:53) Who owns the apps?(1:17:27) Data teams now provide context, not insights(1:18:56) Building with the WisdomAI MCP(1:19:43) Your context is your IP(1:20:48) Closing thoughts: none of that work goes to waste Links from the episode:Meet the Modern Data Team (WisdomAI CDO report)AI Context Engineer (ACE) certificationLive AppsWisdomAI in ChatGPT WorkAvoid AI Writingssot-checkI Paid an AI Agent $8 to Write About its 'Life'Slack Wants to Be the Context Harness for Code | CPO Jaime DeLangheThe AI Framework Era Is Over: Why Context Is the Moat | Jerry Liu Connect with Kapil Chhabra:LinkedInWisdomAIWisdomAI on X Connect with Chain of Thought host Conor Bronsdon:NewsletterTwitter/XLinkedInYouTube More episodes: https://chainofthought.show Thanks to WisdomAI for sponsoring this episode. WisdomAI is the agentic analytics platform for trusted enterprise intelligence: governed context, an analytics harness that makes every answer consistent and verifiable, and Live Apps built from a single prompt. Try Live Apps: https://wisdom.ai/liveapps

Laurie Voss co-founded npm - now head of developer relations at Arize, he argues that engineers will increasingly earn their keep as 'product engineers': understanding what users need and directing AI agents to build it. Full transcript and chapters: https://chainofthought.show/podcast/74-ai-codes-product-engineers-decide-what-to-build-laurie-voss/ One example: a bakery owner who knows how to make a croissant but has no interest in building software. Someone still has to turn that owner's needs into requirements. Laurie sees that work becoming central to product engineering, with cheaper code making software for narrower industries more viable. We discuss where he still sees a need for human code review and operational knowledge, what he would look for in a computer science course if he were starting out today (and what he wouldn't do), and why he compares AI today to the web in 1997. He is optimistic about the technology and skeptical of the valuations, while leaving one question unresolved: how do junior engineers learn the judgment this work demands? We cover: Why the "aha moment" of solving a problem survives even when agents type the codeWhere Laurie still sees a need for human code review, operations, and tacit knowledgeWhether a $15,000 coding bootcamp or a theory-heavy CS degree is still worth itWhy AI in 2026 looks like the web in 1997, and what that says about the bubbleHow AI-generated pull requests burden open source maintainers, and why Laurie expects cheaper code to pressure closed-source business modelsWhy the systems analyst returns as the product engineer, and why that means niche software for bakeries and auto partsWhy Laurie predicts open-model competition and diminishing returns could compress frontier-model marginsHow apprenticeships could help junior engineers develop product judgmentChapters: (0:00) The shift in software jobs(0:44) Why Laurie is optimistic about the code generation explosion(3:06) The aha moment moves from typing code to thinking(5:49) Where agents still need human review and operational knowledge(8:51) Is college still worth it?(9:54) Bootcamps versus theory-heavy CS courses(13:04) We are all product engineers now(15:55) AI is the web in 1997(18:19) Exponential growth, the labs' pause, and npm's ten-year curve(20:07) Barring AGI, AI is a normal technology(21:48) Block's layoffs and companies staying smaller(23:14) What the labor data shows: fewer people, more capital(26:29) Open source as the canary: drowning in AI pull requests(28:11) AI reimplementations and the pressure on software moats(30:26) Personal software and the kill-my-SaaS hackathon(32:32) The bakery and the return of the systems analyst(34:09) Niche software for specific industries(35:36) Bootstrapping and the DevTools opportunity(37:13) What this means for the model companies(38:20) Frontier-model margins and open-model competition(39:53) How the bubble pops: scaling laws and diminishing returns(43:24) Staying private and the trough of disappointment(45:51) Get good at a domain, not the technology(50:44) The missing junior ladder is the question of our time(53:32) Closing thoughts: it's 1997, you can retrain Connect with Laurie Voss: Blog: https://seldo.com/LinkedIn: https://www.linkedin.com/in/seldo/Twitter/X: https://x.com/seldoBluesky: https://bsky.app/profile/seldo.comArize: https://arize.com/Connect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdonMore episodes: https://chainofthought.show Thanks to Svix, presenting sponsor of season four of Chain of Thought. Svix delivers billions of reliable webhooks for startups and the Fortune 500. Get started at https://link.svix.com/cot. Qualified startups get $12,000 in credits, and YC companies get $50,000. Thanks to Walrus, presenting sponsor of season four of Chain of Thought. Walrus Memory gives AI agents portable, verifiable memory that carries context across apps, sessions, and other agents. Get started at https://walrus.xyz/cot Thanks to G2i for sponsoring this episode - for over a decade, they vetted and placed engineers at other companies, from startups to FAANG. Two years ago, they turned that same judgment inward, building their own bench to review RL environments, evals, and training data that models are trained on. Get access: https://fandf.co/3SFxVm6 Thanks to Inngest, presenting sponsor of season four of Chain of Thought. Agents in production run long - they call models and wait on APIs and people. But the longer agents run, the more they break. Inngest handles that with durable execution. You build your agent as steps in TypeScript, Python, or Go. When a step fails, Inngest retries it with exponential backoff, and completed steps are saved and skipped. Try it out: https://inngest.link/cot-pod 7Mdv8UhrombOD7fnCVai

Genspark went from launch to $250 million in ARR in about a year. Along the way it shipped a card-thin meeting recorder, open sourced an office suite that Wen Sang says one engineer prototyped in a week, and started running product triage with agents instead of product managers. Full transcript and chapters: https://chainofthought.show/podcast/73-gensparks-bet-ai-agents-become-the-users-of-software/ Wen's bet is that agents, not people, become the next users of software. Wen Sang is co-founder and COO of Genspark. In this episode he walks through the company's three-layer architecture (models, tools and premium data as the execution layer, a memory layer he calls the second brain, and a collaboration layer called Gen Team), why a meeting note should be the start of work rather than the end of it, the engineering behind the SecondBrain Note, and where he thinks knowledge work goes once agents absorb the busy work. Disclosure: Genspark provided the SecondBrain Note recorder discussed in this episode at no cost. Genspark is not a sponsor of this episode. We cover: Why Genspark builds the self-driving car around the frontier labs' engines, and what that means for people who cannot codeHow Genspark's mixture-of-agents architecture routes work across 70+ models, 150+ in-house tools and paid data setsEvals that grade whether the sales proposal answered the RFP, not whether the model can solve a differential equationWhat a meeting turns into a week later when an agent needs it: proposals, pricing models, research, follow-upsThe SecondBrain Note's microphone array, battery decisions, and consent in a two-party stateWhy GenOffice went open source, and the one-week prototype story behind itHow Genspark runs product feedback triage with agents and no dedicated PMsChapters: (0:00) Cold open (0:32) Geniuses with goldfish memories (3:47) Engines and vehicles: Genspark builds the self-driving car (6:25) Mixture of agents: models, in-house tools and premium data (10:03) Grade the work output, not the intelligence (11:22) A meeting note is where the work starts (13:25) The second brain: Genspark's memory layer (14:45) A thousand recorders, one question for the revenue team (16:33) Execution, memory and collaboration layers (19:19) Ten days in Bora Bora without a laptop (20:07) Gen Team, Slack, and meeting customers where they are (21:57) Agents become the users of software (24:17) Keeping memories current when the deal changes (26:41) Engineering the SecondBrain Note (30:18) The note as an API for the room (31:19) What deserves hardware and what stays software (33:33) Learning hardware supply chains at a two-year-old company (35:05) Why GenOffice went open source (38:01) What knowledge workers do once the busy work is gone (40:07) Building on Genspark with the CLI (42:01) Consent, two-party states and the surveillance line (43:45) Genspark Claw (47:09) Cheaper hardware, deeper integration, and model welfare (51:13) Eighty people and a lot of agents (53:03) What 2027 looks like (59:26) Where to find Wen, and product triage without PMs Connect with Wen Sang: LinkedIn: https://www.linkedin.com/in/wen-sang/Twitter/X: https://x.com/sang_wenGenspark: https://www.genspark.aiGenOffice on GitHub: https://github.com/genspark-ai/genofficeConnect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdonMore episodes: https://chainofthought.show Thanks to Walrus, presenting sponsor of season four of Chain of Thought. Walrus Memory gives AI agents portable, verifiable memory that carries context across apps, sessions, and other agents. Get started at https://walrus.xyz/cot Thanks to Svix, presenting sponsor of season four of Chain of Thought. Svix delivers billions of reliable webhooks for startups and the Fortune 500. Get started at https://link.svix.com/cot. Qualified startups get $12,000 in credits, and YC companies get $50,000. Thanks to Inngest,  presenting sponsor of season four of Chain of Thought.  Agents in production run long - they call models and wait on APIs and people. But the longer agents run, the more they break. Inngest handles that with durable execution. You build your agent as steps in TypeScript, Python, or Go. When a step fails, Inngest retries it with exponential backoff, and completed steps are saved and skipped. Try it out: https://inngest.link/cot-pod Thanks to G2i for sponsoring this episode - for over a decade, they vetted and placed engineers at other companies, from startups to FAANG. Two years ago, they turned that same judgment inward, building their own bench to review RL environments, evals, and training data,that  models are trained on. Get access: https://fandf.co/3SFxVm6

Watch the full conversation on YouTube Full transcript and chapters: https://chainofthought.show/podcast/72-tuning-gpu-performance-with-ai-agents-amds-anush-elangovan-on-rocm-10/ AI agents can keep tuning GPU workloads after you step away from the keyboard. Anush Elangovan, Corporate VP of AI Software at AMD, returns to Chain of Thought to explain how that works with Hyperloom and ROCm 10. Anush and Conor Bronsdon trace the process from installing ROCm through Claude Code or Codex to profiling workloads, finding slow kernels, and testing optimizations while preserving numerical accuracy. Anush shares a Hyperloom run spanning 14,000 models and explains why clear goals and feedback matter when agents are doing the tuning. They also explore what comes next for engineers: keeping skills and frameworks reliable, managing the security and accountability of autonomous agents, and applying AI to the last mile of useful software. We cover: How agents help install ROCm and serve models through natural languageHow Hyperloom profiles workloads and uses LLMs to explore optimizationsThe role of GEAK in tuning kernels while preserving numerical accuracyAnush’s account of optimizing 14,000 models in one passHow software improvements get more performance from existing GPUsKeeping agent skills current and testing across AI frameworksSecurity, accountability, and the next bottlenecks in agent-driven developmentChapters: (0:25) A decade of ROCm, now agent native(3:03) What agentic ROCm looks like in practice(6:17) Installing ROCm then versus now(9:14) An order of magnitude more CI across every framework(10:44) Anush’s workflow: agents and deployment(12:21) Speed is the moat(15:00) Success is a stranger who cannot spell ROCm serving an LLM(17:09) Keeping agent skills from going stale(21:38) Co-designing kernels with the frontier labs(24:06) Hyperloom, GEAK, and 14,000 models in one pass(26:45) Managing autonomous agents: control and liability(32:21) Security at the speed of agent swarms(36:03) ROCm performance gains on the same hardware(37:36) Where enterprises hit walls in production(40:24) Why coding was the right reward function for AI(44:42) Which industries get the next software scale unlock(47:02) The last mile of AI(50:31) Closing thoughts Connect with Anush Elangovan: LinkedIn: https://www.linkedin.com/in/anushelangovan/Twitter/X: https://x.com/AnushElangovanROCm.AI: https://rocm.aiAMD AI blog: https://www.amd.com/en/blogs/by-author/anush-elangovan.htmlAMD AI Developer Program: https://www.amd.com/en/developer/ai-dev-program.htmlConnect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdonMore episodes: https://chainofthought.show Thanks to Svix, presenting sponsor of season four of Chain of Thought. Svix delivers billions of reliable webhooks for startups and the Fortune 500. Get started at https://link.svix.com/cot. Qualified startups get $12,000 in credits, and YC companies get $50,000. Thanks to Walrus, presenting sponsor of season four of Chain of Thought. Walrus Memory gives AI agents portable, verifiable memory that carries context across apps, sessions, and other agents. Get started at https://walrus.xyz/cot

Jaime DeLanghe has spent nine years at Slack turning search, machine learning, and now agents into product. Her team just shipped Slack Code: tag a coding agent like Claude Code, Devin, Codex, or the GitHub agent in a conversation, and it spins up a code channel where everyone in that conversation gets a live development environment, diffs post as artifacts, and the channel winds down when the task is done. Full transcript and chapters: https://chainofthought.show/podcast/71-slack-wants-to-be-the-context-harness-for-code-cpo-jaime-delanghe/ Slack's bet is that AI at work is multiplayer. Agents belong in the channels where teams already work, not in a private chat with one person. Jaime explains why Anthropic pushes so much of its code through Slack, how the channel permission model became the agent context model, and what has to change in engineering culture when the branch is public and the whole team is steering the same agent. The bigger question is whether Slack becomes the context harness where enterprise agents actually run. In this conversation: What happens mechanically when an agent creates a code channel, from authentication to diffs as artifactsWhy engineering at Slack now looks like delegating discrete tasks to agents instead of copy-pasting from a chatHow Slack's channel permission model doubles as the context and access model for agentsWhy Anthropic ships code through Slack: the conversation is where the issue emergesHow culture decides whether a multiplayer coding session converges or splitsWhy solo-terminal coding with an "army of Claudes" reinforces bias, and what social spaces fixSlack as an accidental knowledge management system that ranks recency and engagement over correctness(0:00) Slack as an IDE and a GitHub for your team(0:29) Who is Jaime DeLanghe(1:21) The reaction to the Slack Code launch(5:30) Why coding agents belong in a context-rich environment(6:08) Engineers now manage agents, not copy-paste code(7:24) The permission model: agents get the channel's context(11:44) What happens when a code channel is created(15:00) Why Anthropic pushes so much code through Slack(19:14) Steering one agent with many people: culture decides(24:54) Slackbot, skills, and MCPs: agents go where the work is(30:53) The solo terminal vs. agents in social spaces(33:53) Org charts and ownership when agents join the team(39:33) Learning loops and shared agent memory(42:39) Citations, recency, and accidental knowledge management(46:50) Context bloat and multi-pass search for agents(50:01) How Jaime uses Slackbot as CPO(52:38) Slack Code is V1 of multiplayer AI Connect with Jaime DeLanghe: LinkedIn: https://www.linkedin.com/in/jaime-delanghe-aba59b1a/Slack Code announcement: https://slack.com/blog/news/slack-code-channels-for-agentsIntroducing Slack Code (Salesforce): https://www.salesforce.com/introducing-slack-code/Connect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdonMore episodes: https://chainofthought.show Thanks to Walrus Memory, presenting sponsor of season four of Chain of Thought. Walrus Memory gives AI agents portable, verifiable memory that carries context across apps, sessions, and other agents. Get started at https://walrus.xyz/cot. Thanks to Svix, presenting sponsor of season four of Chain of Thought. Svix delivers billions of reliable webhooks for startups and the Fortune 500. Get started at https://link.svix.com/cot. Qualified startups get $12,000 in credits, and YC companies get $50,000.

Tormod Ree puts 11 or more cameras and dozens of microphones into a single meeting room, then runs computer vision on all of it to figure out who is present, who is talking, and who is looking at whom. He is the chief product and engineering officer at Neat, the Zoom-backed video hardware company. Full transcript and chapters: https://chainofthought.show/podcast/70-11-cameras-dozens-of-mics-the-ai-that-reads-the-room/ Before Neat, Tormod co-founded AVA, a computer vision security company Motorola acquired, and spent close to eight years at Cisco running the Spark Board. He explains how Neat turns a room into a system that directs the meeting instead of just framing whoever talks, why almost all of the AI has to run at the edge, and how the company builds computer vision models without ever collecting a customer's audio or video. In this conversation: Why the real advantage is the harness that combines signals from different detectors, not the models themselvesHow Neat draws a hard line: no customer audio or video ever trains its modelsThe "captain" device that orchestrates a room full of cameras and mics, directing from metadata rather than re-running inference on every streamWhy the meeting "director" is still deterministic today, and what changes when it becomes a trained modelBuilding AI under a five-year device lifetime and phone-class computeAgentic fleet management, an MCP server, and what Tormod calls "agentic healing"How Neat gets its own non-technical teams building with AI(0:00) Reading the room: 11 cameras, dozens of mics(0:27) Who is Tormod Ree(2:07) Turning a meeting room into a system that directs itself(3:39) What it takes to actually read a room(5:13) Why the media path has to run at the edge(6:27) Open models, in-house models, and the harness that matters(7:51) The data problem when you can't touch customer meetings(11:09) The captain device: distributing compute across the room(15:38) Two users: the people in the room and the IT admin(17:00) Agentic management and "agentic healing" for device fleets(20:12) Why a meeting device has to stay useful for five years(23:44) How agentic workflows evolve on an open platform(25:53) When the meeting director becomes a trained model(28:54) Building AI under five-year, phone-class hardware limits(37:26) Where silicon caps what you can run locally(39:18) AI pendants and other form factors(41:48) How Neat adopts AI across its own teams(49:20) Non-technical teams building their own MCPs(50:27) Where Neat is headed Connect with Tormod Ree: LinkedIn: https://www.linkedin.com/in/toree/Neat: https://neat.noConnect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdonMore episodes: https://chainofthought.show Thanks to Svix, presenting sponsor of season four of Chain of Thought. Svix delivers billions of reliable webhooks for startups and the Fortune 500. Get started at https://link.svix.com/cot. Qualified startups get $12,000 in credits, and YC companies get $50,000. Thanks to Walrus Memory, presenting sponsor of season four of Chain of Thought. Walrus Memory gives AI agents portable, verifiable memory that carries context across apps, sessions, and other agents. Get started at https://walrus.xyz/cot.

Behind Thomson, the new legal AI model from Thomson Reuters, is a $40 million investment in people, compute, and evaluation methods. The final training run cost just $450,000. CTO Joel Hron, whose teams build Westlaw, Practical Law, and CoCounsel for millions of professionals in more than 100 countries, joined us for the launch to break down why the 175-year-old company chose to own its model layer instead of solely renting frontier intelligence. Full transcript and chapters: https://chainofthought.show/podcast/69-thomson-1-the-new-40m-legal-ai-model-thomson-reuters-joel-hron/ We cover: Why Thomson Reuters trained its own model instead of relying only on Claude, GPT, or GeminiThe compute, data, and expertise flywheel behind the Thomson modelHow rebuilding CoCounsel around agent-native tools took one-shot accuracy from 25% to over 70%The rent-versus-buy case for owning model weights and compounding expert feedback over timeThe dangers of AI inaccuracies in legal workCitation ledgers, deep research, and verifying legal work with no ground-truth oracleJoel's advice to CTOs weighing open models and training on their own dataChapters: (0:00) Cold open: a $40M model and 25% to 70%(0:27) Why Thomson Reuters built the Thomson model(2:46) From information services to an AI company(5:14) The flywheel: compute, data, and expertise(8:21) The oldest company to ship a model?(9:47) Training for users without catastrophic forgetting(14:37) Continuous pre-training on Westlaw and Checkpoint(15:29) Fine-tuning, DPO, and agentic reinforcement learning(17:37) Rebuilding CoCounsel: 25% to 70% overnight(21:48) Capturing expert judgment: own versus rent the model(28:59) Managing lawyer time and protecting customer IP(31:45) Eval results and avoiding catastrophic forgetting(34:34) Tabular analysis and legal deep research(36:26) Benchmarks, Harvey, and frontier comparisons(39:26) Verifying legal work with no ground-truth oracle(41:54) Citation ledgers and the hallucinations that matter(45:41) Rebuilding the platform and the Trust in AI Alliance(48:59) Advice to CTOs on open models and owning intelligence(51:31) The compounding flywheel and what comes next Connect with Joel Hron: LinkedIn: https://www.linkedin.com/in/joel-hron-90a3421a/Thomson Reuters AI: https://www.thomsonreuters.com/en/artificial-intelligenceThe Thomson model and next-gen CoCounsel Legal: https://www.thomsonreuters.com/en/press-releases/2026/august/thomson-reuters-launches-next-generation-of-cocounsel-legal-the-ai-ecosystem-built-for-legal-professionalsConnect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdonMore episodes: https://chainofthought.show Our sponsors: Thanks to Svix, presenting sponsor of season four of Chain of Thought. Svix delivers billions of reliable webhooks for startups and the Fortune 500. Get started at https://link.svix.com/cot. Qualified startups get $12,000 in credits, and YC companies get $50,000. Thanks to Walrus Memory, presenting sponsor of season four of Chain of Thought. Walrus Memory gives AI agents portable, verifiable memory that carries context across apps, sessions, and other agents. Get started at https://walrus.xyz/cot.

Attackers used to take months, sometimes 270 days, to weaponize a disclosed vulnerability. Now it happens in weeks, minutes if the incentive is there, and independent reports from Mandiant and CrowdStrike show the average time to exploit has gone negative. Full transcript and chapters: https://chainofthought.show/podcast/68-time-to-exploit-is-negative-ai-broke-the-patch-cycle-dan-lorenc/ Dan Lorenc's conclusion: finding flaws is no longer the hard part. Fixing them first is. Dan is the co-founder and CEO of Chainguard. Before that he spent years at Google building the backbone of software supply chain security and created Sigstore. In June his team launched Athena, a coalition of more than two dozen companies including JPMorgan, Cloudflare, Cisco, and Kyndryl, built for the era where AI finds vulnerabilities faster than maintainers can patch them. Last month alone it processed more than 40,000 AI-discovered findings. In this conversation: Why the average time to exploit went negative, and what collapsed the fat tail of never-exploited bugsHow models chain "low severity" flaws into working exploits, like Project Zero's zero-click iPhone takeoverInside Athena: what happens between a member submitting a finding and a fix landing upstreamWhy 40,000 findings is not 40,000 CVEs: validation, dedup, and vulnerability archaeologyFuzzing outpaced patching for a decade, and AI is the first tool that speeds up the fixing sideThe Log4j thought exercise for solo maintainers and enterprise CISOs alikeFork economics, "state your intentions," and defense in depth for agent infrastructureChapters: (0:00) Cold open: how time to exploit goes negative(0:31) The 20-year assumption that just died(3:21) What a negative time to exploit actually means(6:04) Two new realities: attacks democratized, more bugs than anyone knew(8:55) Chaining tiny flaws: the Project Zero iPhone story(11:26) Why fixing AI-found vulnerabilities takes a coalition(13:27) 40,000 findings in one month: submission to upstream fix(18:06) The agentic pipeline: as few human eyes as possible(19:03) Fuzzing outpaced patching for a decade(20:50) The Log4j thought exercise for maintainers and CISOs(23:49) When no maintainer answers: the new economics of forking(27:40) Deleting dangerous code to slow the treadmill(29:35) How security kills entire vulnerability classes(31:08) Agent infrastructure: defense in depth or nothing(34:35) Regulation: maintainer liability, frontier labs, DC's busy year(37:01) Open model economics(39:55) Gas Town, multiclaude, and going back to normie(42:47) Why Dan turns the AI memory system off(45:53) Closing: still the most fun time to build software Connect with Dan Lorenc: LinkedIn: https://www.linkedin.com/in/danlorenc/X: https://x.com/lorenc_danChainguard: https://www.chainguard.devAthena: https://www.chainguard.dev/athenaConnect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdonMore episodes: https://chainofthought.show Thanks to Svix, presenting sponsor of season four of Chain of Thought. Svix delivers billions of reliable webhooks for startups and the Fortune 500. Get started at https://link.svix.com/cot. Qualified startups get $12,000 in credits, and YC companies get $50,000. Thanks to Walrus, presenting sponsor of season four of Chain of Thought. Walrus Memory gives AI agents portable, verifiable memory that carries context across apps, sessions, and other agents. Get started at https://walrus.xyz/cot.

Agents are like teenagers: profoundly intelligent, extremely resourceful, no fear of consequence, and missing the judgment to know right from wrong at all times. That's how Cisco President and Chief Product Officer Jeetu Patel thinks about securing AI agents, and it's why he says static allow/block rules are already obsolete. Agents are smart enough to route around them. Full transcript and chapters: https://chainofthought.show/podcast/67-parenting-your-ai-agents-for-best-results-ciscos-jeetu-patel/ Jeetu returns to Chain of Thought to map cyber's third phase: an agent trust platform where security and observability fuse into one discipline. He explains how Cisco Cloud Control spins up a digital twin to test every agent-recommended fix before it touches production, why Cisco moved from unlimited tokens to rationing them like headcount, and why the gap between people who are fluent with AI and people who aren't is now a 10x differential, not 10%. He also makes the contrarian case that AI will create more jobs than it destroys  - and of course, we talk infrastructure for this new era.  We cover: Why agents need dynamic runtime guardrails instead of static allow/block rulesThe agent trust platform: how security and observability are fusing into one disciplineAgentic ops in Cisco Cloud Control: ambient agents, digital twins, and human-in-the-loopHow Cisco rations AI tokens the way it rations headcountThe intelligence, cost, and control trade-offs behind open vs closed modelsAction control vs access control: what rights you actually grant an agentWhy every automation step creates a new human bottleneck, and more jobsChapters: (0:00) Agents are like teenagers: cold open(0:25) Welcome back Jeetu Patel(1:18) Open weight vs closed models(2:26) Intelligence, cost, and control: the model trade-off triangle(10:13) Shrinking model half-life and the economics of frontier training(11:38) Why token costs still outrun token value(14:15) Build your own evals and route intelligently(17:11) Rationing tokens like headcount at Cisco(19:54) Agentic ops: ambient agents and digital twins in Cisco Cloud Control(23:01) When to take the human out of the loop(25:13) Cyber's third phase: the agent trust platform(28:17) Parenting AI agents: dynamic boundary conditions, not static rules(31:10) LiveProtect and baking security into the network fabric(34:40) Action control vs access control for agents(36:42) What the industry is getting wrong(37:48) The case for AI creating more jobs and the 10x fluency gap(42:36) Career paths, entry-level hiring, and upskilling at scale(45:40) Closing thoughts Connect with Jeetu Patel: LinkedIn: https://www.linkedin.com/in/jeetupatel/X: https://x.com/jpatel41Cisco: https://www.cisco.comConnect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdonMore episodes: https://chainofthought.show Thanks to Svix, presenting sponsor of season four of Chain of Thought. Svix delivers billions of reliable webhooks for startups and the Fortune 500. Get started at https://link.svix.com/cot. Qualified startups get $12,000 in credits, and YC companies get $50,000.

Jitender Aswani was customer zero for Presto at Meta, where a billion daily active users generated queries that took hours to return. He watched that drop to minutes, scaled the same technology at Netflix across 300 million subscribers, and now runs engineering and security at Starburst, the $3.35 billion platform built on Trino. Full transcript and chapters: https://chainofthought.show/podcast/66-data-federation-not-centralization-is-what-enterprise-ai-needs/ His argument: every enterprise AI project that stalls is fighting the same hidden battle. The agents can query the model fine. They just can't reach the data. The average enterprise runs 52 to 200 data sources, and a decade of moving all of it into one lake produced ETL debt, governance problems, and pipelines that break whenever a SaaS vendor adds a column. Federation is the only model that scales with entropy. We cover: Why Presto changed what Meta could experiment on, and how that compounded product velocityWhat broke when Jitender took the same technology to enterprises running 52 to 200 data sourcesWhy centralization stopped working once data grew faster than the ability to move itWhat happened to Starburst's query volume the day they shipped an MCP serverThe FinOps agent that fired queries for 30 minutes against data it never hadHow AIDA turns ad hoc analysis into workflows using skills and MCP serversWhy a context graph is different from a knowledge graph, and why ontology decides agent accuracy(0:00) Enterprises run on 52 to 200 data sources(0:25) Intro(2:18) Customer zero for Presto at Meta(9:50) Scaling to trillions of events at Netflix(15:11) Taking Trino from Silicon Valley to 10,000 enterprises(20:24) The 2011 research that predicted conversational analytics(28:54) Why centralization can't scale with entropy(32:26) The agent query explosion and what MCP did to volume(41:44) Inside AIDA, Starburst's conversational analytics product(46:56) Context graphs versus knowledge graphs(51:17) Where to follow Jitender's work Connect with Jitender Aswani: LinkedIn: https://www.linkedin.com/in/jitenderaswani/Starburst: https://www.starburst.io/Connect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdonMore episodes: https://chainofthought.show

Charles Guillemet is CTO of Ledger and the founder of the Donjon, Ledger's internal offensive security lab whose job is to break the company's own products before attackers do. He spent a decade in cryptography and hardware security before Ledger, including designing secure integrated circuits. Full transcript and chapters: https://chainofthought.show/podcast/65-you-cant-secure-an-ai-agent-with-software/ His argument is blunt: you cannot secure an AI agent with software alone. As agents start moving real money, API keys and trust scopes leave no physical verification layer, and Charles makes the case that hardware has to sit in the loop. This one turned into a wide-ranging thought piece (and some debate) on what the agentic economy actually looks like, and how to stay safe inside it. We cover: Why Charles thinks "securing an AI agent" with software permissions and API keys is a false promiseThe economic asymmetry between attackers and defenders, and how AI is collapsing itHow a policy engine plus a hardware-enforced signature can delegate rights to an agent safelyWhy Charles thinks the agentic economy settles on blockchain rails over Visa and MastercardSecure elements, HSMs, and zero-knowledge proofs as execution-integrity guaranteesHow Ledger uses hardware authorization internally for passkeys, signed releases, and multisigA practical way to classify assets by threat model and match security to value(0:00) Why securing an AI agent in software alone is impossible(0:30) Delegating execution power inside your security perimeter(2:28) The attack-defense asymmetry AI is erasing(6:00) The alignment problem and delegating rights to agents(9:24) Policy engines, intents, and hardware-enforced signatures(13:19) From developer experience to agent experience(15:12) Secure elements, HSMs, and execution integrity(20:00) Zero-knowledge proofs, proving without revealing(27:24) Convincing the skeptics on agent-driven payments(34:49) Why Ledger bet on dedicated hardware(36:15) Hardware as a determinism layer for agents(38:52) How Ledger uses hardware authorization internally(43:42) Classifying assets by threat model(46:55) When attack and defense become symmetric(48:44) Deepfakes, voice cloning, and the scam wave(50:04) Closing thoughts on staying safe in the agentic economy Connect with Charles Guillemet: LinkedIn: https://www.linkedin.com/in/charles-guillemet/Twitter/X: https://x.com/P3b7_Ledger: https://www.ledger.comConnect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdonMore episodes: https://chainofthought.show

Jiaona Zhang(JZ) is the Chief Product Officer at Laurel, where the team runs its own product on itself to see exactly where AI helps and where it doesn't. Before Laurel, JZ built products at Airbnb, Dropbox, Webflow, and Linktree, and she has taught product management at Stanford for nearly a decade. Full transcript and chapters: https://chainofthought.show/podcast/64-stop-token-maxxing-find-where-ai-actually-pays-off-jiaona-zhang/ Companies are spending billions on AI tooling, but most still can't say where it returns time or revenue. Jiaona breaks down how to get that visibility, why blanket AI mandates backfire, and what it takes to re-architect a team so anyone can ship. Her argument is simple: stop token maxing and start measuring time back. We cover: Why most organizations can't see where AI is actually working, and how Laurel uses time data to fix itThe token max trap that "use AI everywhere" mandates create, and how to drive efficient use insteadWhy former managers make the best operators of agent fleetsHow Laurel lets PMs, designers, and customer success ship features end to endThe bottom-up plus top-down playbook for re-architecting a team around AIWhy technology moats are falling away while brand and data moats endureLaurel's bet on returning time to people instead of replacing them(0:00) The token max trap(1:47) Why companies can't see where AI is working(5:03) What Laurel does: turning time into data(8:53) Agents as an extension of the workforce(13:43) Why former managers make the best AI users(18:23) Lean teams and shipping end to end(22:29) Enabling non-engineers to ship features(28:30) Re-architecting teams: bottom-up and top-down(32:09) Keeping your professional identity as AI shifts work(38:53) The context layer is the new race(42:06) Fundamentals plus tinkering: how to learn(48:45) Brand and data moats when tech moats fall away(54:31) Laurel's movement: returning time to people Connect with Jiaona Zhang(JZ): LinkedIn: https://www.linkedin.com/in/jiaona/Laurel: https://www.laurel.ai/JZ's Linktree: https://linktr.ee/jzConnect with Chain of Thought host Conor Bronsdon: Newsletter: https://newsletter.chainofthought.show/Twitter/X: https://x.com/ConorBronsdonLinkedIn: https://www.linkedin.com/in/conorbronsdon/YouTube: https://www.youtube.com/@ConorBronsdonMore episodes: https://chainofthought.show