Kwadwo Adu

I started in IT in 2004, moved into digital design and growth, and over time began building products, teams, and companies. Today, I run PLYO Lab, an applied-AI advisory and venture studio, where we use AI to build experiences, agentic systems, and products for businesses.

Originally from Copenhagen, Denmark, now based in Barcelona, Spain, I've spent years across startups, scaleups, and corporate venture and innovation labs including SPACE10 (IKEA), Prehype, dkbn, and Trendsales, as well as co-founding and building bootstrapped or VC-funded startups like Bloggers Delight, Achoo, and Gray Matter.

I run an Applied AI community of around 2,000 people, host hackathons, and write a newsletter about practical AI, product, and business.

I tinker with real estate projects here in Spain with my wife.

Now

PLYO LabFounder · advisory & venture studio2018–

From AI curiosity to production, then build and ship. 10+ ventures for LEGO Ventures, COTY, Dimexon, Danske Spil, Norlys, Prehype.

Where I worked

Gray MatterFounder & CEO2023–25

Applied-AI studio. 13 products shipped, a 15-person team, ~€500K raised (Morph VC).

SPACE10Head of Digital · IKEA R&D2015–18

IKEA's research and design lab. Global digital innovation, AI / AR / MR.

AchooCOO & Co-Founder2014–15

Influencer-marketing SaaS.

Bloggers DelightCo-Founder2010–14

Nordic creator platform, 1M+ monthly users.

ByNightOperations Lead → Head of Operations2004–09

From Operations Lead to Head of Operations.

Side missions

buildingLocaleFlow · AI translations for Shopify stores, always in sync
buildingShipfeed · every AI release in one feed, operated by AI agents
buildingSignkit · email-signature SaaS, operated by AI agents
buildingBookey · coworking & office SaaS, operated by AI agents

Community

writingNewsletter · practical AI, product, and business
hostingApplied AI Meetup · Barcelona's AI-builder meetup, 15+ editions, ~2,000 people
hostingHKTN.dev · Professional hackathons, clients bring a real problem, builders ship

Open source

open sourcemodelfit · benchmark LLMs on your own codebase · ★5