Devinson Peña
I build products and operating systems that make complex systems easier to trust and use.
I’m based in Zurich and work across product, partnerships, tokenization, onchain products, and practical AI. The thread through all of it is the same: technically capable systems only become useful once the workflow around them is clear enough for real people to trust and use.
Selected work
ChopDot
A Polkadot-native group expense app built around one practical problem: shared money only works when the group understands what is happening.
Started at a global Polkadot hackathon, won second prize, and now anchors closeout proof on Polkadot Hub through an EVM smart contract.
YourTurn
A booking and resale product for studios, therapists, and coaching-led services where a reservation behaves more like a pass than a disposable slot.
Built at ETHGlobal Cannes 2026 and won the Hedera "No Solidity Allowed" track.
CAS Blockchain at HSLU
Formal blockchain work in one of the more serious Swiss rooms for tokenization, market structure, and legal-commercial discussion.
Included a DePIN tokenomics paper, presentation, and a stronger public signal of serious blockchain credibility in Switzerland.
DePIN / tokenization research
Research focused on what happens when incentive systems face real pressure instead of living inside optimistic assumptions.
Used Onocoy as an anchor case to test tokenomics under slower growth, weaker retention, and real network stress, then turned that work into a live stress-test dashboard.
AI / agent systems
A practical thread focused on reviewable agents, browser workflows with approval gates, and research flows tied to evidence.
Built as public proof through practical-agent-systems rather than keeping the work trapped in private notes or demos.
What I work on
Tokenization and digital asset products
Where legal structure, product workflow, and client understanding all have to hold together in practice.
Wallets, coordination, and trust-heavy flows
Especially where shared state, money movement, or permissions become hard to understand once more than one person is involved.
Partnerships, GTM, and adoption work
When a product is capable but the story, onboarding, or internal alignment still needs tightening before it can move.
Practical AI
Where automation and agent tooling need to become useful inside real work rather than stay as demos.
Background
Before the current Web3 and AI work, I spent more than fifteen years across product programs, GTM, onboarding, client delivery, and operating systems in environments where strong tools still needed better structure, translation, and trust around how they were used.
Explore
If any of that sounds useful, the cleanest next step is a short call or a direct note.