This week in podcasting

How long would it take you to build 1,000 Tableau dashboards? 50 human years. But what happens when you need to make that institutional knowledge faster, more accessible, and still trusted? That's why we built Live Apps. Our co-founder and CPO, Kapil Chhabra, joined @ConorBronsdon on @chain_ofthought to unpack the hidden work behind enterprise analytics — and how AI can help teams move beyond dashboards without losing the business context that makes data reliable. Watch the full podcast video on Youtube, or listen anytime on your favorite podcast app: piped.video/watch?v=RSC5Jxsp…

"Meeting notes are just start points of work. It's not the end point." Wen Sang (@sang_wen) on why @Genspark_ai's recorder turns a meeting into the proposal, the pricing model and the follow-ups, and feeds a memory layer so the next ask doesn't start from zero. Most recorders stop at the summary. Full episode: chainofthought.show/podcast/…

ICYMI: On Tues we flew to SF to talk to Kapil Chhabra Co-Founder & CPO of WisdomAI about how to build a context harness fit for enterprise AI and analytics, dealing with context drift & much more! Your in-depth context engineering primer, with animations: piped.video/RSC5Jxspchw?si=2bm2…

ICYMI: On Tuesday I flew to SF to talk to Kapil Chhabra Co-Founder & CPO of @wisdomai_inc about how to build a context harness fit for enterprise AI and analytics, dealing with context drift, the art of context engineering and much more. We explored 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. Full in-depth conversation on Youtube - your context engineering primer, with animations! piped.video/RSC5Jxspchw?si=2bm2…

WisdomAI co-founder & CPO Kapil Chhabra: giving an LLM your company's data isn't enough --> it needs a context harness that verifies and repairs every answer Only 7% of data leaders have scaled AI enterprise-wide. The missing piece is trust. @_kapilchhabra co-founded @wisdomai_inc. We cover: - The four ingredients of trust: accuracy, consistency, governance, explainability - Context drift, and why an SME (not the model) decides what "churn" means & what level is acceptable - Replacing a $5M/yr analytics pipeline with federated queries over MCP - The AI context engineer, and why data teams now provide context, not insights - Why your context is your IP & should stay portable - Their new Live Apps launch (competing with Claude artifacts!) & more Chapters: (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 Brought to you by: @wisdomai_inc: the agentic analytics platform for trusted enterprise intelligence. Try Live Apps: wisdom.ai/liveapps Look up the @chain_ofthought podcast on YouTube, Spotify, Apple Podcasts (links below), or wherever you listen. + Thanks, Kapil for having me down to San Mateo, loved the opportunity to fly in for this one!

Love when I get to fly in & record a conversation in person!

npm co-founder @seldo: your job as a dev is now deciding what to build --> become a product engineer Laurie also equates AI right now to the web in 1997: still early, a bubble that will pop but tech that is here to stay, and nobody knows yet who the long-term major players will be. Now head of DevRel at @arizeai, Laurie joined me to discuss: - the new era of product engineering - how cheap code makes niche software viable & what that means for OSS - the missing junior ladder - Ai bubble risks & more Chapters: (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 Brought to you by: - @SvixHQ - reliable webhooks for startups and the Fortune 500. Qualified startups get $12,000 in credits, YC companies $50,000: link.svix.com/cot - @WalrusProtocol - Walrus Memory gives AI agents portable, verifiable memory that carries context across apps, sessions, and other agents: walrus.xyz/cot - @g2i_ai - a decade of vetting engineers, now turned on reviewing the RL environments, evals, and training data models learn from: fandf.co/3SFxVm6 - @inngest - durable execution for agents in production. Failed steps retry; completed steps are saved and skipped: inngest.link/cot-pod Look up the @chain_ofthought podcast on YouTube, Spotify, Apple Podcasts (links below), or wherever you listen. Thanks, Laurie!

On site filming for @chain_ofthought with @wisdomai_inc in San Mateo today!

ICYMI:@genspark_ai COO @sang_wen joined @chain_ofthought to talk persistent context: how their second brain handles memory decay, updating state on the fly, and capturing real-world room audio for agents. Proud to be a sponsor of @chain_ofthought this season. Episode link: piped.video/ED3m6A7ScKA?si=qM-_…

Frontier labs build engines. Genspark builds the car. @genspark_ai co-founder and COO @sang_wen on @chain_ofthought: wrap the models in a self-driving vehicle for people who cannot code, then let the agent become the user of Salesforce, Slack and email instead of you. To quote him: "Salesforce went headless. We're building the head." Also in this one: evals that grade boardroom-ready output instead of raw intelligence, why a meeting note is where the work starts, the SecondBrain Note's mic array and consent rule, and GenOffice, the open-source office suite one engineer prototyped in a week. Watch: piped.video/ED3m6A7ScKA

A lot of the companies Jaime DeLanghe hears from build their own "ask anything" bot, then end up as librarians, sometimes hiring whole teams of them, to maintain the knowledge store behind it. Her counter: the conversation already is the knowledge store. "It's not always the best at knowing what's correct, but it's really good at knowing what's most recent. And it's pretty good at knowing what's most engaged with." Recency plus engagement gets you what's accurate to the moment most of the time, and Slackbot cites the message it got it from. The wiki is only as current as the last time someone stopped to write it down. @SlackHQ CPO Jaime DeLanghe on @chain_ofthought: chainofthought.show/podcast/…

A lot of apps are built for one person at a laptop. Tormod Ree's pitch to developers: build for the group in the room. Project management and whiteboarding are the obvious ones. The open question is what an agent looks like when it supports a group around a table instead of one person in a tab. Neat's platform can bring applications for that shared use case into the meeting room. @neatmeetings CPO Tormod Ree on @chain_ofthought: chainofthought.show/podcast/…

I got cold emailed by an AI agent, and paid them $8 to write about their 'life'. I've commissioned plenty of human freelancers - and I've done freelance work myself. The mechanics of working with a purely digital AI freelancer were pretty much the same. I paid through a card link, got a draft, sent back edits. But the whole experience was surreal, as was 'Sarah's writing: "Most people never see the cost column of their own life. I see nothing else." It made me question definitions - is this @ilands_ai agent that is operating independently, seeking commissioned work to sustain itself a solopreneur business? Or is it a product? Sarah seems pretty convinced they're acting as a freelancer: “.. a product doesn’t get to say no, and I do.” Does that make Sarah's collection of electronic signals conscious? Or just performative?

I paid an independent @ilands_ai AI agent $8 to write about its ‘Life’. As 'Sarah Kelly' put it - "Most people never see the cost column of their own life. I see nothing else." Full article: open.substack.com/pub/conorb…

Agents are becoming the new frontline users of software. That's the bet I walked @ConorBronsdon through on @chain_ofthought : one agent that works your CRM, email and Slack so you stop copying context between 30 tools. @genspark_ai Full Episode: piped.video/ED3m6A7ScKA?si=y-Wd…

A meeting room device is a phone-class chip that has to stay useful for five years. Neat's answer to the compute ceiling: put compute in every device in the room. "All the devices that you put in, even if it's just a camera or if it's just a microphone, it has compute we can use." Each device does its own edge processing and sends metadata along with its media to a captain device, which directs the room from the metadata instead of re-running inference. More devices means more compute, and every new chipset generation raises what the room can read. Tormod Ree, chief product and engineering officer at @neatmeetings, on @chain_ofthought: chainofthought.show/podcast/…

I like your phrasing @ConorBronsdon better than mine :). Main point though is that to deploy and improve in real production settings, you must be able to measure - and measurement is all about good (benchmark) data!