Built This Week

This week on Built This Week we sit down with Anis Bennaceur, co-founder and CEO of Attention, to demo the AI-native Revenue Operating System that automates CRM updates, sales coaching, and pipeline forecasting after every call. Sam also demos a Monte Carlo forecast dashboard he built on top of Attention's open source coaching stack.

We cover:

• What Attention is and how it automates sales operations end to end
• Sam's live Monte Carlo forecast dashboard built on Attention's open source stack
• How AI listens to every call and updates your CRM automatically
• Why filling the CRM is the painkiller every sales team actually needs
• Real-time coaching scorecards and how reps get scored during the call
• Pipeline forecasting and how AI catches deal risk before you do
• How to manage LLM token costs at scale without killing developer productivity
• Why Anis personally approves every token budget overrage at Attention
• Anthropic filing for an IPO and what it means for the AI race
• Why Uber capping developer token spend is the wrong move
• Agents vs deterministic workflows — the honest truth about what actually works in production
• Why the companies letting their teams run free with AI are pulling ahead

If you are in sales, building a revenue operations stack, or want to understand where AI is taking enterprise software, this episode is for you.

Find Anis and Attention:

Website: attention.com

LinkedIn: linkedin.com/in/anis-bennaceur

Twitter: x.com/anisbennaceur1

Company LinkedIn: linkedin.com/company/attentiontech

Company Twitter: x.com/tryattention

TIMESTAMPS

[00:00] Intro

[01:03] Meet Anis — CEO and co-founder of Attention

[01:50] Sam demos his Monte Carlo forecast built on Attention's open source stack

[04:51] What Attention actually does — the core product

[06:30] Painkiller vs vitamin — why CRM accuracy is the real problem

[07:45] Real-time coaching scorecards and call scoring

[09:20] Pipeline forecasting and deal risk detection

[11:00] Who is using Attention and why

[12:30] Automating work for reps, managers, leaders, and RevOps

[14:00] News — Anthropic filing for an IPO

[15:10] Uber CEO capping developer token spend

[17:00] How Attention tracks LLM cost per client, per feature

[19:05] Claude token caps and how Anis approves overrages personally

[20:30] The token budget chaos — tracing where the spend went

[22:30] Agents vs deterministic workflows — what actually works

[24:34] Why workflows beat agents in most production scenarios

[25:10] How to find Anis and Attention

[25:30] Wrap up

New episodes every Friday at BuiltThisWeek.com

#AI #Sales #CRM #BuildWithAI #Anthropic #AIAgents #RevOps #BuiltThisWeek #Attention

This week on Built This Week we sit down with Wiley Jones, CEO and co-founder of Doss, to demo an AI-native operations platform built for mid-market companies managing the flow of goods, dollars, and data. He incorporated the company a month before ChatGPT existed.

We cover:

• What Doss is and why he calls it an Adaptive Resource Platform not an ERP
• A live demo of the platform including inventory management, procurement, and order management
• How Doss lets companies build their own data model instead of forcing them into the shape of the software
• DOSBOT — an AI agent that can self-introspect the entire system and answer complex operational questions
• Why switching costs in enterprise software are only relevant if you are talking to the wrong customer
• How to identify the 5 to 10 percent of the market that actually wants to move
• The ICP — physical product companies doing 20 million to a few hundred million in revenue
• News — Robinhood launching AI agents to trade stocks and whether that is innovation or gambling
• Whether AI trading bots can actually generate alpha or just raise everyone's tide equally

If you are building in enterprise software, working in operations, or want to understand where AI is taking business systems, this episode is for you.

This week on Built This Week we sit down with Adir, CEO and co-founder of Autonomy AI, to demo the platform helping enterprise teams build, update, and ship product changes directly on top of their existing codebase.

No slides. Just a live walkthrough of the real product.

We cover:
What Autonomy AI is building
How non-technical teams can work directly with existing codebases
Why the handoff between product, design, and engineering is still broken
Building new product views from a simple prompt
Connecting AI-generated work to real APIs and pull requests
Design mode, Figma-style editing, and mobile responsiveness
Why companies are using less Figma and Jira
Who is actually buying AI product-building tools
Karpathy joining Anthropic and what it means for the space
Google’s new AI agents and the future of search

If you are building with AI, managing product teams, or trying to understand how software development workflows are changing, this episode is for you.

This week on Built This Week we sit down with Karim, founder and CEO of Breeze (heybreez.ai), to demo the enterprise voice agent platform built for companies that actually need to run AI at scale. No slides. No prep. Just a live walkthrough of the real product.

We cover:
What Breeze is and how it differs from other voice AI platforms

• A live demo of the platform built from a blank screen
• Why multi-agent architecture matters for enterprise compliance
• How to prevent voice agents from being jailbroken
• The latency and model selection decisions that make or break production
• The operational layer of voice AI that everyone is ignoring
• Phone groups, smart routing, and running 10,000 calls a day
• Who is actually buying enterprise voice AI and why
• Breeze's partnership model and expansion into MENA and South America
• Anthropic raising at a $900 billion valuation and what it means
• Whether AI is replacing junior developers and what comes next

If you are building with AI, scaling voice agents, or trying to understand where enterprise AI is heading next, this episode is for you.

Find Karim and Breeze:

Website: heybreez.ai

LinkedIn: linkedin.com/in/kmalhas

Twitter: x.com/kmalhas_

TIMESTAMPS

[00:00] Intro

[01:05] Meet Karim — founder and CEO of Breeze

[02:12] Live demo — building a voice agent from scratch

[07:34] Multi-agent architecture explained

[11:55] Latency vs quality tradeoff in voice AI

[12:56] OTP security and jailbreak prevention

[14:13] The operational layer nobody is building

[16:41] Version control, compliance, and phone groups

[18:43] Who is the ideal customer for Breeze

[21:49] Partnership model and global expansion

[22:55] News — Anthropic raising at $900B valuation

[24:28] Will AI replace junior developers

[26:05] Building from the Middle East — why it matters

[28:23] Where to find Karim and Breeze

[28:52] Wrap up

New episodes every Friday at BuiltThisWeek.com

#VoiceAI #AIAgents #Enterprise #Anthropic #BuildWithAI #BuiltThisWeek #AIStartup #heybreez

This week on Built This Week we sit down with Ramses Alcaide, founder of Neural, to demo a noninvasive brain computer interface that tracks your focus in real time through a pair of headphones. No surgery. No implants. Just data.

We cover:

• What a brain computer interface actually is and how it works
• A live demo of focus tracking during the episode
• A browser plugin that adjusts podcast speed based on your brain activity
• How the technology detects brain fatigue before you feel it
• Gaming, medical, and sports applications
• How AI finally unlocked BCI for consumer devices after 40 years in labs
• The science behind ice baths affecting men and women differently
• Why kids are bypassing age verification AI with a fake mustache
• Why AI security cameras are still failing at basic common sense
• The edge compute problem nobody in consumer hardware wants to talk about
• Why your brain signature might be the future of identity verification

If you are building with AI, interested in wearables, or want to understand what brain computer interfaces actually are today, this episode is for you.

TIMESTAMPS

[00:00] Intro

[00:44] Meet Ramses — what is a brain computer interface

[01:40] How the headphones track focus in real time

[02:28] Live focus tracking demo on the podcast

[03:00] Browser plugin that adjusts podcast speed to your brain

[04:35] How to use biofeedback to stay focused

[07:02] Brain health tracking — cognitive strain and brain age

[07:33] Gaming use case — overclocking your brain with HP

[08:16] ER doctors and high stakes focus applications

[09:08] How AI finally brought BCI out of the lab

[10:01] Origin story — PhD, family tragedy, US Army backing

[11:45] How to find Ramses and the product

[13:09] Ice bath experiment — men vs women brain data

[14:45] News — AI security cameras calling dogs bears

[15:13] Why edge AI for consumer hardware is brutally hard

[16:28] Brain data and camera AI share the same constraint

[17:08] Kids bypassing age verification with a fake mustache

[19:23] Brain signature as the future of identity verification

[20:03] Wrap up

New episodes every Friday at BuiltThisWeek.com

#BCI #AITools #Neuroscience #BuiltThisWeek #BrainComputerInterface #AI #Wearables #EdgeAI

This week on Built This Week, we break down one of the most interesting new AI product launches in recent memory: Claude Design.
No demos. No fluff. Just what happens when AI starts replacing traditional design workflows.
We cover:
 • What Claude Design is and how it works
 • Creating ad campaigns, decks, and full product redesigns with simple prompts
 • Why it could become a serious competitor to tools like Figma
 • How teams are exporting AI designs directly into production code
 • The rumored xAI / Cursor deal and what it means for the coding race
 • ChatGPT Images 2.0 and whether it lives up to the hype
 • Why Google might be quieter now—but still dangerous long term
If you're building with AI, working in design, or trying to understand where creative tools are heading next, this episode is for you.
⏱ TIMESTAMPS
[00:00] Intro
 [00:45] Claude Design overview
 [01:50] First impressions after using Claude Design
 [03:00] How the interface works
 [04:20] Building decks, ads, and redesigns with prompts
 [06:10] Creating ad campaigns for Hip Train
 [07:45] Exporting projects, sharing, and production handoff
 [10:15] Full internal app redesign with AI
 [12:45] Is Claude Design a Figma killer?
 [13:00] xAI / Cursor acquisition rumors
 [16:15] ChatGPT Images 2.0 reactions
 [18:30] Why AI is still in the early innings
 [21:40] Google’s new TPUs and staying in the race
 [22:40] Wrap up & what’s next for Built This Week

Links
BuiltThisWeek.com
New episodes every Friday

Jordan Metzner
https://x.com/mrjmetz

Sam Nadler
https://x.com/Gravino05

This week on Built This Week, we break down how AI helped a non-technical teammate build a real internal product that now helps teams move faster across the company.

No buzzwords. No fake use cases. Just real AI in production.

We cover:
 • The AI tool that automates meeting follow-up work
 • How transcripts become tickets, reports, and dashboards
 • Why internal AI products are becoming a huge advantage
 • How companies can train non-technical teams to build
 • What happens when everyone can create software
 • Why the next wave of AI is about empowerment
If you're serious about using AI to improve your business, this episode is for you.

⏱ TIMESTAMPS
(00:00) Intro
 (00:32) Welcome back
 (00:40) Guest introduction
 (01:28) Inside the Radar tool
 (02:36) Solving workflow bottlenecks with AI
 (03:23) Instant task generation from meetings
 (04:45) Smarter project visibility with dashboards
 (05:56) Real productivity gains
 (07:00) From personal tool to company product
 (08:08) Future roadmap
 (09:24) AI-generated business reviews
 (10:19) Building an AI-first culture
 (11:30) Teaching non-technical teams
 (12:49) Real examples across departments
 (13:57) Why this changes work forever
 (15:06) News segment
 (17:44) Closing thoughts

🎙 HOST INFO
Hosted by Jordan Metzner and Sam Nadler
 Co-founders of Ryz Labs
We build AI-native companies and tools used by startups, enterprises, and investors.

🔗 CTA + LINKS
Subscribe for weekly breakdowns of real AI builds and what actually matters
 New episodes every Friday

Follow along:
 YouTube: Built This Week
 Spotify: Built This Week
 Apple: Built This Week

This week on Built This Week, we break down a real AI tool we built that’s already saving days of work in production.

No demos. No fluff. Just how AI is actually being used inside a real business.

We cover:
 • The internal tool that replaced complex spreadsheets and cut turnaround time in half
 • How we generate client-ready presentations instantly with AI
 • Why tool selection matters more than ever in the agentic era
 • Google AI Studio and how we use it to prototype fast
 • Anthropic’s unreleased model and what it means for AI safety
 • Meta’s latest push into AI and why competition is heating up

If you're building with AI or thinking about how to apply it inside your company, this episode is for you.

⏱ TIMESTAMPS

00:00 Intro
 00:40 What we’re covering this week
 01:30 The problem with planning large offsites
 02:28 How you lose money without perfect cost visibility
 03:23 The AI tool we built (Offsite estimator)
 04:29 Hidden costs AI catches that humans miss
 05:36 From spreadsheets to automated workflows
 06:04 Instant client presentations with AI
 07:19 Cutting turnaround time from 10 days to 3
 08:09 Tech stack behind the tool (Codex, Supabase, React, AWS)
 08:58 Real customer impact and results
 10:05 What we’re building next (automation + client portal)

10:40 Google AI Studio deep dive
 11:14 How we actually use it for prototyping
 12:55 Image, music, and video generation tools
 14:48 When to use which AI tool

15:33 The real framework for choosing AI tools
 16:45 Anthropic’s unreleased model
 17:54 Why it might be a security risk
 18:32 Who should control powerful AI

19:45 Meta’s new AI push
 21:10 Why competition is accelerating

22:30 Wrap up

🎙 HOST INFO

Hosted by Jordan Metzner and Sam Nadler
Co-founders of Ryz Labs

We build AI-native companies and tools used by startups, enterprises, and investors.

Training gets the headlines.

Inference is where the money is.

In Episode 37 of Built This Week, we sit down with Mitesh, CEO of Positron AI, to break down one of the biggest bottlenecks in AI today: inference infrastructure.

While the world focuses on trillion-parameter models and frontier labs, the real constraint isn’t intelligence — it’s memory, bandwidth, energy, and cost.

We cover:

• Why inference is where 90% of AI spend happens
 • The memory wall problem in large models
 • Why GPUs weren’t designed for text generation
 • How Positron is building terabyte-plus memory chips
 • The economics of 10 trillion parameter models
 • Why memory bandwidth utilization matters
 • Why CPUs are suddenly back in demand
 • The difference between speed-optimized and cost-optimized AI systems
 • The slider bar future of AI infrastructure

We also dive into:

• OpenAI’s $122B valuation
 • Anthropic vs OpenAI secondary market dynamics
 • Why Nvidia isn’t going anywhere
 • Why commodity memory might beat premium stacks in certain use cases
 • The rise of agentic workflows and what that means for compute

If you care about the future of AI, silicon, infrastructure, or trillion-dollar companies — this episode is for you.

New episodes every Friday.

⏱ TIMESTAMPS

(0:00) Why inference is the real AI bottleneck
 (2:00) What Positron AI is building
 (4:30) The memory problem in trillion-parameter models
 (6:30) Why GPUs struggle with inference economics
 (9:00) Energy, bandwidth, and supply chain constraints
 (12:00) Memory capacity vs memory speed tradeoffs
 (16:00) The “slider bar” model of AI infrastructure
 (18:30) OpenAI’s $122B valuation discussion
 (21:00) Anthropic vs OpenAI secondary markets
 (23:30) CPUs making a comeback
 (26:00) Agentic workflows and compute demand explosion
 (28:00) Closing thoughts on AI infrastructure

Our recruiters are not video editors.

But now they can cut highlight reels in minutes.

In Episode 36 of Built This Week, we break down a tool we built internally at Ryz Labs that lets our recruiting team generate polished candidate highlight videos without touching Premiere, Final Cut, or CapCut.

The problem:

When presenting candidates to clients, resumes are standard.
 But seeing a candidate speak for 60 seconds changes everything.

The issue was speed.
 Editing sizzle reels required our video team, added delays, and was not scalable.

So we built a highlight reel generator powered by:

• EntreVista AI interview transcripts
 • Claude and Codex for clip selection
 • Remotion for video rendering via code
 • AWS S3 for instant share links

The system automatically:
 • Analyzes transcripts
 • Identifies high signal clips
 • Groups them by communication, role fit, and personality
 • Allows light manual adjustments
 • Renders a branded video in 5 to 10 minutes

No editing experience required.

Then we dive into Remotion and why “video as code” is one of the most underrated AI enabled workflows right now.

Finally, we discuss the growing cost of AI usage inside organizations:
 • Token spend management
 • Surprise AI bills
 • Model access guardrails
 • Productivity vs cost tradeoffs

AI is democratizing building.

But it is also introducing a new management layer.

New episodes every Friday.

⏱ TIMESTAMPS

(0:00) The problem: recruiters are not video editors
 (0:25) Welcome to Episode 36
 (1:20) Why highlight reels improve candidate selection
 (2:30) The scalability issue with manual video editing
 (3:30) Demo: AI Highlight Reel Builder
 (4:15) How transcripts power automatic clip selection
 (5:00) Communication, role fit, personality grouping
 (6:10) Manual adjustments for recruiters
 (7:00) Rendering time and infrastructure challenges
 (8:00) Final sizzle reel output demo
 (9:00) How it was built with Codex
 (10:00) What is Remotion
 (11:30) Video editing as code explained
 (12:30) Other Remotion use cases: product trailers, documentation videos
 (13:45) Democratizing creative production
 (14:30) AI token costs inside organizations
 (15:15) Surprise AI bills and infrastructure lessons
 (16:30) Managing model access across teams
 (17:30) Productivity vs spend tradeoffs
 (18:15) Closing thoughts

🔗 LINKS

Built This Week
 New episodes every Friday
 https://builtthisweek.com

Jordan Metzner
 https://x.com/mrjmetz

Sam Nadler
 https://x.com/Gravino05