Katrina Laszlo
Product design for the agent era
I design products where humans set direction and agents execute. 10 years of product, growth, and monetization across startups and enterprise.
Design isn't just the UI.
It's the pricing, the onboarding, the API, the agent behavior, the service. When agents are users too, design moves beyond the screen. The experience is the system.
How I think
- Human defines the minimum. Agent executes.
- The best products give humans control over structure and success criteria. Everything else is delegated. This is the design pattern for everything that comes next.
- UX without UI.
- Service design has been designing experiences without screens since the 1980s. Call center journeys, service blueprints, moments of trust. Nobody is applying these methods to agents. That's the gap.
- Data and design are two sides of the same coin.
- Both are about understanding user behavior. Design observes how people use a product. Data measures it. The best product decisions come from treating them as one discipline. Hypotheses, testing, rapid iteration. That's the loop whether you're running an experiment or redesigning an experience.
Experience
2025 – Present
Co-Founder, CEO at Tanso
Monetization infrastructure for AI and B2B SaaS.
2025
AI Product Manager at Artisan
2024 – 2025
Head of Product at Chipper
Consumer fintech. Debt elimination through intelligent automation.
2021 – 2024
Senior PM, Web & Apps at People Data Labs
Data-as-a-service. Backed by Founders Fund and Craft Ventures.
2019 – 2021
Corporate Innovation at 500 Global
Venture capital. Startup acceleration and corporate innovation.
2017 – 2019
Product Consultant at EY
Innovation Lab. Building and launching financial services ventures.
2014 – 2016
Founder of Nosh Delivery
Student-to-student food delivery at Georgetown.
Georgetown University McDonough School of Business
BS Business Administration. International Business, Political Economics, Marketing.
Greek, English, Spanish
Projects
Open source tools I've built.
zero-to-ship
An open-source toolkit for product people who are ready to build.
agentlens
See what your AI coding agents store on your machine. Scan memory, token costs, session transcripts.
observe
Open source AI cost observability. See which customers and features are unprofitable.
buildnext
Point it at your customer interviews. Get back the themes you'd find in two weeks.
dx-test
Inspector-level scrutiny of the end-user experience, automated.