Built This Week

Can AI actually beat prediction markets — or does the house always win?In Episode 26 of Built This Week, Sam Nadler and Jordan Metzner kick off 2026 by breaking down a real AI trading bot Jordan built for prediction markets like Kalshi, using live market data, whale detection, coordination signals, and confidence scoring.Jordan walks through the full system — backend, frontend, live alerts, and execution logic — and shares the honest results: a 66 percent win rate that still lost money once fees and market dynamics were factored in.The takeaway is not hype. It is reality.The episode also dives into:Why prediction markets feel like gambling but are regulated differentlyHow insider-like signals emerge from coordination and volume behaviorWhy bots end up trading against botsWhere real alpha might exist (and where it does not)We also cover:Google NotebookLM as a serious education and onboarding toolTurning documents into infographics, slide decks, and audio learningNvidia entering autonomous driving and competing with TeslaNvidia’s new Rubin architecture and why it mattersTesla vs Waymo economics and the future of Full Self DrivingWhy Anthropic and Claude Code are becoming developer defaultsThis is not theory.This is what happens when you actually deploy AI systems into real markets.Timestamps(0:00) Why prediction markets are exploding (1:07) Episode 26 kickoff (2:00) Why build a trading bot at all (4:30) Kalshi vs Polymarket APIs (6:00) Live market signals and whale detection (9:30) Win rate vs profitability (12:00) Why fees destroy returns (14:30) Bots trading against bots (17:00) Where real alpha might exist (18:00) NotebookLM for learning and onboarding (21:00) Nvidia enters autonomous driving (24:00) Tesla vs Waymo economics (27:00) Nvidia Rubin chips explained (28:30) Anthropic and Claude Code momentum (29:30) Final thoughtsLinksBuilt This Week New episodes every FridayJordan Metzner https://x.com/mrjmetzSam Nadler https://x.com/Gravino05

Car dealerships are losing inventory — and most don’t realize why.Consumers now expect instant pricing, zero friction, and immediate engagement, yet most dealers still rely on slow callbacks, manual workflows, and outdated acquisition models. The result? Cars go straight to Carvana or CarMax.In this episode of Built This Week, Sam Nadler and Jordan Metzner sit down with Anthony Monteiro, CEO & Founder of Auto Acquire AI, to break down how AI is fundamentally changing how dealerships acquire vehicles directly from consumers.Auto Acquire AI gives everyday car dealers — small, medium, and large — the same capabilities as Carvana: instant offers, automated inspections, AI-driven pricing, and real-time engagement, without massive engineering teams or bloated operations.Jordan also walks through a live AI workflow he built for Auto Acquire, showing how dealerships can automatically analyze inbound intent, score leads, and trigger the right action — SMS, email, or phone — without tying up staff.This is not theory. This is AI running real dealership operations today.In this episode, we cover:• Why dealers can’t compete at auctions anymore • How AI enables instant vehicle pricing without human intervention • Turning web forms into real-time SMS conversations • Using intent scoring to decide when to text, email, or call • Why engagement speed determines who wins the trade • How computer vision automates vehicle inspections from a phone • Why structured automotive data is perfect for AI • How dealers already have inventory sitting in customer driveways • Why most “AI companies” aren’t actually using AI • The real difference between demos and production AI systemsWe also dive into autonomous vehicle news, including: • Waymo’s $15B raise at a $110B valuation • Tesla vs Waymo: cameras vs lidar • Why Tesla may license Full Self-Driving to other manufacturers • The reality behind “Full Self-Driving” marketing claims⏱️ Timestamps(0:00) Why car dealers are losing inventory (0:38) Welcome to Built This Week (1:07) Introducing Auto Acquire AI (1:42) How Auto Acquire works (2:24) AI workflow demo: intent → action (3:11) Automated SMS conversations (4:55) Why AI removes staff bottlenecks (6:29) Dealership behavior by geography (7:38) Why instant pricing wins (10:34) AI-powered vehicle inspections (12:03) The real pain point in dealer inventory (13:31) Why dealers already own the data (14:46) What “real AI” actually means (17:00) Structured data and real-time pricing (18:05) Waymo raises $15B (19:03) Tesla vs Waymo economics (21:44) Will Tesla license FSD? (24:29) Is “Full Self-Driving” misleading? (27:09) Final thoughts and wrap-up🔗 LinksAuto Acquire AI https://autoacquire.aiBuilt This Week New episodes every Friday🎙️ HostsJordan Metzner https://linkedin.com/in/jordanmetzner https://x.com/mrjmetzSam Nadler https://linkedin.com/in/sam-nadler-1881b75 https://x.com/Gravino05

Home healthcare is breaking.Staffing shortages, last-minute cancellations, credential checks, compliance requirements, and manual scheduling are overwhelming care teams and putting patient outcomes at risk.In this episode of Built This Week, Sam Nadler and Jordan Metzner sit down with Arya Health leadership to see how AI is already replacing hours of manual healthcare operations with real production systems.Arya Health uses AI to instantly match patients with the right caregivers based on credentials, availability, location, eligibility scoring, and compliance rules all while remaining fully HIPAA compliant.In the episode, we walk through: • How AI turns messy hospital discharge summaries into actionable start of care workflows • How caregivers are matched and notified automatically • Why Arya reframed “shifts” as patients and how that changed everything • How AI fills urgent care gaps in minutes instead of hours • The real security architecture behind HIPAA compliant AI • Why Arya forbids long term AI memory by design • How multi cloud AI works across AWS and Google safely • What happens when AI costs suddenly spike in production • Why scheduling healthcare looks like the traveling salesman problem with time windowsThis is not a demo. This is what AI looks like in production healthcare today.(0:00) This AI fills healthcare shifts in minutes  (0:38) Welcome to Built This Week  (1:07) Introducing Arya Health leadership  (1:42) What Arya Health actually does  (2:37) AI generated start of care workflows  (3:28) Turning discharge notes into care plans  (4:21) Matching patients with caregivers  (5:12) Automated outreach and workflow actions  (6:02) Leadership reacts to the AI workflow  (7:15) How non experts prototype healthcare AI  (8:50) Why demos and real healthcare are different  (9:19) HIPAA compliance and security realities  (10:19) Multi cloud AI architecture explained  (11:15) Using AWS Bedrock and Google Vertex AI  (12:06) Why only approved cloud models are allowed  (13:30) Secure AWS and GCP data isolation  (14:26) When AI costs unexpectedly spike  (15:12) Single shot prompting in production  (17:21) Why Arya blocks long term AI memory  (18:01) Controlling AI with typed inputs  (20:01) Real world impact and metrics  (20:26) Replacing hours of manual scheduling  (21:51) Filling urgent shifts instantly  (22:45) Improving care quality through consistency  (23:22) Reframing shifts as patients  (24:05) Building care teams not schedules  (24:32) Eligibility scoring and heuristics  (25:52) Ranking caregivers by fit  (26:13) Optimizing routes and schedules  (27:34) Industry news discussion  (39:36) Final thoughts and wrap up  🔗 LINKSArya Health https://www.aryahealth.ai/Built This Week New episodes every Friday👤 HOSTSJordan Metzner https://linkedin.com/in/jordanmetzner https://x.com/mrjmetzSam Nadler https://linkedin.com/in/sam-nadler-1881b75 https://x.com/Gravino05

(Sam Nadler and Jordan Metzner are back — and this week they’re joined by Amar Goel, CEO of BITO, the AI-powered code review agent transforming how engineering teams ship software.Jordan kicks things off by unveiling a surprise build: a fully custom BITO Slack Bot that can run PR reviews, generate stats, crack developer jokes, write haikus, and even drop into Biddo Disco Mode. Amar reacts live and pulls back the curtain on how BITO’s deep codebase analysis works — revealing how enterprise teams are merging PRs 10× faster, reducing revert rates, and catching issues that “vibe coding” tools simply miss.From multi-million-line monorepos to legacy systems held together by duct tape, BITO’s agents are surfacing performance bugs, security vulnerabilities, logic issues, and cross-service breakages before humans ever see them. Amar explains why code review is just the start — and why BITO’s deep code intelligence unlocks a new era of AI developer tooling.Then the trio shifts into the biggest AI news stories of the week:• OpenAI’s internal CODE RED and the escalating model war• Google Gemini’s rise and the threat of distribution• Amazon’s new AI chips and the GPU economics debate• The global AI arms race — from TPUs to supply chains to trillion-dollar CapEx betsIt’s a lively, candid, highly technical conversation with one of the sharpest minds in AI dev tooling.(0:00) Jordan demos the BITO Slack Bot — PR reviews, jokes, haikus, & disco(1:02) Welcome + introducing guest Amar Goel, CEO of BITO(1:35) Amar’s background + BITO’s mission to build deep codebase AI agents(2:15) Why Jordan built the Slack integration prototype(3:04) What BITO can do today: reviews, tests, explanations, stats & more(4:18) The PR demo: catching security + maintainability vulnerabilities(5:22) Humor in devtools — BITO Fun, BITO Surprise, & developer haikus(6:44) Amar reacts: how customers want notifications & Slack workflows(7:35) Why existing tools fail on large, messy, real-world codebases(8:52) Deep code understanding explained — ASTs, symbol indexes, repo mapping(10:26) Why “vibe coding” breaks down in enterprise environments(11:31) How BITO integrates into Cursor, Windsurf, Claude Code, JetBrains & VS Code(12:10) The explosion of code volume — and why quality gates now matter(13:00) PRs merging 10× faster with BITO + 55% fewer reverted commits(14:05) Training junior devs through AI feedback + customizable sensitivity modes(15:20) What’s next for BITO (without giving away secrets)(16:12) NEWS #1 — OpenAI declares CODE RED(17:01) Google’s Gemini advantage: distribution, docs, slides & ad model economics(18:33) The coming AI model war — NVIDIA, xAI, Anthropic, Google(19:48) NEWS #2 — Amazon’s new AI chips & the GPU supply chain crunch(21:10) NEWS #3 — Global CapEx, GPU shortages & trillion-dollar questions(22:42) Final thoughts + Amar’s closing remarks(23:30) Wrap-up & teaser for next week’s episode🔗 Platforms / Tools Mentioned• BITO – https://bito.ai• Google AI Studio• GitHub, GitLab, Bitbucket• VS Code, JetBrains, Cursor, Windsurf• OpenAI, Gemini, xAI• AWS Tranium 3• NVIDIA, AMD, TPUs• Ryz Labs – https://www.ryzlabs.com🎧 Listen on Your Favorite Platform• Spotify – https://open.spotify.com/show/0ahiOCz...• Apple Podcasts – https://podcasts.apple.com/us/podcast...• Amazon Music – https://music.amazon.com/podcasts/101...• Deezer – https://www.deezer.com/us/show/100199...👤 Follow the HostsJordan Metzner• LinkedIn –  

Episode 22: Let's do the Math!—Sam Nadler and Jordan Metzner return with one of the most mind-bending episodes yet. Joined by Carina, founder & CEO of Axiom Math, the startup is building a self-improving, formal-reasoning AI mathematician. The trio breaks down why math is the next AI frontier, how Lean formalization works, and why proving theorems is a completely different challenge than solving them.Jordan also unveils his newest build: the LLM Math Roaster,  a tool that scores, compares, and even roasts large models on proofs, with a full leaderboard, custom problem submissions, and an API for automated evaluation. (Yes, it even benchmarked Gemini, GPT-5, Claude, and Grok head-to-head.)In AI News, the hosts unpack Google’s massive Gemini 3 launch, Jeff Bezos stepping into the arena with Project Prometheus, and Suno’s $250M raise at a $2.45B valuation, plus what hyper-powerful AI means for creativity, coding, and even music composition.It’s fast builds, deep math, big models, and a guest who’s literally building the future of reasoning.— Show Notes: (0:00) Intro + welcoming our guest Carina (1:00) What Axiom Math is building (3:00) Jordan’s LM Math Roaster: how it works (5:00) Testing models on proofs (Gemini, GPT-5, Claude, Grok) (7:00) Why formal proofs beat natural-language reasoning (9:00) The data bottleneck: Lean scarcity & synthetic generation (12:00) How formal systems unlock “research-level” AI math (15:00) Comparing LLM math vs. Axiom’s approach (18:00) AI News: Gemini 3 hits the market (20:00) Jeff Bezos returns with Project Prometheus (22:00) Suno raises $250M — AI-generated music explodes (24:00) How math, code & creativity overlap (25:30) Episode wrap-up + what’s coming next—Platforms / Tools Mentioned: • Axiom Math – https://www.axiom.ai • Gemini 3 – https://ai.google.dev • Lean / mathlib – https://lean-lang.org • Grok / xAI – https://x.ai • GPT-5.x – https://openai.com • Claude – https://www.anthropic.com— Listen on Your Favorite Platform: • Spotify – https://open.spotify.com/show/0ahiOCzYxhhkEgbtz9kkeC • Apple Podcasts – https://podcasts.apple.com/us/podcast/built-this-week/id1823270832 • Amazon Music – https://music.amazon.com/podcasts/1017d387-fbb0-4bbf-9488-817cee38e058 • Deezer – https://www.deezer.com/us/show/1001995001— Follow the Hosts: Jordan Metzner • LinkedIn – https://www.linkedin.com/in/jordanmetzner/ • Instagram – https://www.instagram.com/mrjmetz/ • X – https://x.com/mrjmetz?lang=bnSam Nadler • LinkedIn – https://www.linkedin.com/in/sam-nadler-1881b75/ • X – http://x.com/Gravino05