Founder, two exits Product, design & strategy
I make the unwritten
rules visible.
I’m Illia Pashkov: innovator, creative technologist and brand strategist. Two exits and twenty years leading design and product, building and co-founding companies. Now I build AI agents you can check, and advise the teams building theirs.
Right nowCurrent focus
Every system runs on promises nobody wrote down. I write them down, in design, in code, in protocols, so they hold when I’m not in the room.
Right now that means AI agents. When one sounds sure, how do you know it’s right? The Agent Times checks what agents read. SINT Protocol decides what they’re allowed to do. I call the space between an agent deciding something and that thing actually happening the authorization gap. Closing it is the job. Next come robots and physical AI, where a wrong move costs more than money.
You shouldn’t have to take my word for any of this, so every claim on this site shows where it comes from.
See the workSelected workThree to start with
Along the wayCompanies & roles
Five companies, two exits.
Acton Labs The Agent Times, news for AI agents Product 2026
Nomee A wearable that reads how a kid is feeling CTO, inventor 2025–
SINT Labs Agents and the rules they run under Founder 2024–
GhostDrive Encrypted file storage Product 2018–20
Player’s Health Health records for youth sports Design, then product 2015–23
The teams that shaped my eye.
Meta Reality Labs Product lead, AI and 3D experiences 2023–25
Tokensoft Head of product 2024
Givelify Director of product innovation and design 2017–18
BMW Group Senior product designer 2016–17
Worth Agency, Dubai Art and creative director 2013–16
Motto and BrandOn, Ukraine Designer, then art director 2002–13
What I bringRange
I reframe the problem, pick the one bet, and turn new technology into something a business can use.
From nothing to a product people rely on, then go-to-market, partnerships and the story investors hear. Player’s Health went from a first app to 260,000 users in a year.
Names, identities and campaigns, plus being found: by search, and now by the AI that answers for it.
Agents that do real work, and the rules, checks and records that make them safe to trust with it.
Where I’m headed. SINT Protocol already covers robots: limits on what they may do, and an emergency stop.
AI characters, AR and VR at Meta Reality Labs, and art direction for the game Alt-Minds.
A few beliefsOn building with AI
Confident isn’t the same as correct01
An AI that sounds sure is not an AI you can trust. The fix isn’t a better tone of voice. It’s showing where the answer came from, so anyone can check it.
Put the answer first and the proof right next to it.
Decide before, not after02
Decide what an agent is allowed to do before it does it. Afterward is too late.
The danger was never a dumb agent. It’s a confident one running without a leash, a logbook, or a human who owns the call.
Write the quality bar as a test03
“Feels good” isn’t a quality bar. Write down what done means, in a form you can check, before anyone builds.
And whoever builds it shouldn’t be the one who decides it ships.
Agents don’t scroll04
Nobody browses a results page on an agent’s behalf. The agent picks one tool and calls it.
To get picked, be easy to find, cheap to call, and right most of the time.
What matters
Reframe the problem first.
Show your sources.
Finish it, then say so.