![Cover Image for Building AI-Native Startups [005]](https://images.lumacdn.com/cdn-cgi/image/format=auto,fit=cover,dpr=2,background=white,quality=75,width=400,height=400/uploads/hh/92921602-f86e-4a3b-9ce0-5798177c417c.png)
Building AI-Native Startups [005]
Online workshop for AI-native builders.
No slides. No theory. Only live demos.
Founders and operators screen-share the workflows actually running their companies: the agents, the harnesses, the org structures, and the parts that break. Real screens, real terminals, in real time.
What it is
A 90-minute live workshop. Three builders each get 15 to 20 minutes to demo one thing they genuinely run day to day, followed by live Q&A. We run these every two weeks.
Who it's for
Founders, solopreneurs and operators rebuilding how they and their companies work around AI. If you are shipping agents in production, or working out what to build next, this is the room.
What's inside
3 × 15-20 min screen-share demos of real, in-production work
Live Q&A after each one
Recording shared with everyone who registers
Why attend live
The demos get published afterwards. The Q&A does not, and that is where the specifics come out: what it cost, what broke, what they tried first and threw away.
Public, free, no sponsors, no ads, no pitches.
SPEAKERS
Karl Wirth — CEO & Co-founder, Nimbalyst
Demo: Visual system for managing agents
Create a feature plan and mockup with an agent.
Review and refine the plan with a PM and developer.
Run parallel coding sessions and review the implementation.
Ship the feature and update the docs and website.
Karl Wirth is the CEO/co-founder of Nimbalyst, the open-source visual workspace for building with coding agents. He previously was co-founder and CEO of Evergage, the leading real-time ML personalization company acquired by Salesforce. At Salesforce, he grew Evergage into the $250M/year Salesforce Personalization business. Karl was CPO and CTO at Treasure Data, and earlier in his career worked at Red Hat and RSA Security. He graduated from Harvard with a degree in Physics.
Nimbalyst · GitHub · X · LinkedIn
Cris Lenta — Founder, LifeSim
Demo: Automated feedback to new features flow
Turn player bug reports and suggestions into tasks.
Check each task against the codebase and roadmap.
Trigger AI coding sessions and review the results.
Generate a daily briefing from viral social content.
Cris Lenta is the founder of LifeSim, an AI-native gaming company building an open-world life simulator with emergent worlds, backed by Founders Inc and Paul Bragiel (Roblox, Unity, Uber, Niantic, Supercell). He co-authored two AI research papers on how neural networks can self-replicate through backpropagation (MIT Press / Artificial Life Journal, IEEE Xplore), and studied CS and Statistics at LMU Munich. His first startup was acqui-hired.
LifeSim · crislenta.com · X · GitHub · LinkedIn
AJ Gleser — CEO & Co-founder, Ergoly
Demo: Running customer success with AI workers
See an org chart with human teammates and AI workers.
Watch a worker post customer digests and flag stuck accounts.
Teach the worker in Slack with memory separated by client.
Debug blocked runs and mistaken claims about what the worker can do.
AJ Gleser is co-founder and CEO of Ergoly, a Seattle company that turns a service firm's know-how into AI workers that live in its Slack, one sealed worker per client. Before Ergoly he ran a 24-person digital product agency, built new ventures for Fortune 1000 companies such as Airbus and Goodyear at Mach49, and ran more than 300 customer interviews across six continents. Ergoly runs on Ergoly: AJ and co-founder Dan Moore manage their team and a fleet of AI workers from inside Slack.
Ergoly · gleser.ai · LinkedIn
Host: Stepan Gershuni
cyber•Fund · AI-Native Monastery accelerator.
LinkedIn · X · cyber•Fund
About cyber•Fund and the Monastery
cyber•Fund has been investing since 2014 and backs AI-native founders. The Monastery is our accelerator for teams building AI-native from day one: $2M uncapped, 5%, a three-month program, built on the idea that focus is a founder's scarcest resource. We will say how to apply at the end of the session.