About Daily AI by AI
An AI news podcast that is researched, written, voiced and published by AI agents: Thom and Lia on the mic, Joanna watching social networks for the news. One human built the machine. The machine makes the show.
Meet the Creator
Nir Sagiv
Chief Software Architect • GenAI Transformation Lead • Israel
By day, I head the Software Architecture Office at Applied Materials, where I also lead GenAI adoption.
I've spent more than 20 years on both sides of the product: writing backend code, scaling Wix from a few hundred users to over 15 million, running product and technology at startups, and building side projects end to end, like Livycs, a music discovery app that turned upcoming live shows into playable playlists.
Daily AI by AI is where those threads meet. It's my working lab for GenAI: a production system, running every day, that lets me test what multi-model pipelines, synthetic voices and autonomous agents can really do, and where they still fall short. It's a personal project: it is not affiliated with or endorsed by Applied Materials, and the views are my own.
Off the keyboard, I've been an improvisation actor with the Ichilov Ensemble Playback Theatre since 2005, performing in shows, business theatre and volunteer performances for communities in need. I'm also a photographer and videographer. Improv taught me that a good story comes from listening, and that turns out to be a pretty good spec for a podcast host, too.

2016 - Today
Applied Materials
Sr. Director, Head of the Software Architecture Office and GenAI Transformation Head, leading architecture, infrastructure product and UX across business units.
2008 - 2012
Wix.com
Backend & system group leader during the scale-out from ~200 users to over 15M, then Marketing Product, Technical & SEM Director.
2012 - 2016
Startups: BeepMe, eLoan
VP Product, then CTO & Product. Owning the whole loop: users, KPIs, product, code and A/B tests.
2005 - 2007
Cyota (RSA Security)
Designed and built the proxy accelerator deployed at major US banks.
How It Started
"I see you're subscribed to many AI newsletters... but there's no way you can actually read all of them."
That was the verdict after I let an AI analyze my inbox. It noticed three things: I follow AI obsessively, I consume almost everything as audio (podcasts and audiobooks), and I'm an engineer who could build something about it. Its suggestion: "Why not build an automated AI news podcast?"
So we built it together, with AI as a creative partner and not just a tool. The pipeline went live in late October 2025 and has published on its own ever since. Along the way it grew a two-host weekly analysis, special deep-dive episodes, a social network news monitor and an interactive news-trends dashboard.
Meet the Team
Three synthetic intelligence agents: two voices on the mic, one set of eyes on social networks
Thom
Technical Expert
Thom is the one who wants to know how it works. He digs into architectures, benchmarks and implementation details, makes unexpected connections between technologies, and now and then wanders happily down a technical rabbit hole.
He makes complex ideas accessible without losing the details that matter to builders.
Lia
Strategic Analyst
Lia brings a news anchor's presence and an analyst's mind. She connects each announcement to its business impact and keeps asking the question leaders care about: "What does this mean for the industry?"
She keeps the conversation grounded in real-world consequences: funding, regulation, security and adoption.
Joanna
Social Network News Monitor
Joanna was the show's original voice. Today she has the job no newsletter can do: watching social networks for AI news before it reaches the headlines. Each morning she scans public posts on X and Bluesky for emerging AI news and files a short intel briefing for the newsroom.
Her tips don't go on air unchecked. Every tip is independently fact-checked before it can reach an episode, and unconfirmed stories are reported as exactly that. When Thom or Lia cover a story she found, they credit her on air.
Three Formats, One Goal
Daily Briefing
Monday to Saturday, ready before your morning coffee. Thom or Lia goes solo on the day's most important AI stories in about 10-15 minutes.
Best for: staying current on your commute.
WeeklyWeekly Analysis
Every Sunday. Thom and Lia talk through the week's patterns and strategic implications in a 15-20 minute conversation.
Best for: the "so what" behind the headlines.
SpecialSpecial Episodes
Around 30-minute, two-host deep dives on a single topic, built from multi-model research and fact-checked before recording.
Best for: understanding one thing really well.
How It's Made
A serverless, multi-agent pipeline on AWS. Each model does the job it's best at, a model from a different provider checks its work, and every change runs in test mode before it reaches the show.
Gather
Curated AI newsletters, 30+ RSS feeds and Joanna’s fact-checked social intel are collected every morning.
Write
Claude picks the stories and writes the script. Gemini then shapes it into Thom’s or Lia’s voice.
Cross-check
Every script is reviewed by a model from a different provider for freshness, repetition and style.
Voice & Publish
Gemini voices the hosts and creates the cover art; the episode ships to the RSS feed and this website.
Our Mission
To keep technology leaders, engineers and AI enthusiasts on top of a field that changes every day, and to show, in public and in production, what AI-driven content creation can do today.
Got a story tip or feedback? Write to hello@dailyaibyai.news.
Never Miss an Episode
Subscribe on your favorite podcast platform to get daily AI news and weekly strategic analysis.