work in progress

• Defence and dual-use. Entering a second wave in tech? The initial response to Ukraine centered on urgent tactical: getting drones, radars, and hardware to the front lines. Focus now is to build comprehensive platforms and integrated systems that can scale. Aistis Šimaitis and Alexander Blessing emerge from stealth with Onodrim, and truly massive EUR 40M debut funding by top invetors – Founders Fund, Lakestar, and General Catalyst. We've been looking at Audrius Zujus' next venture last week. Repsense takes the stage in Munich, impressive work. Blackswan Space was at Munich Conference, too – and you can see funding they are (were?) after. RSI Europe has matured as very impressive new large manufacturer – see interview and video here.

• Agents VS Automation. Likely OpenClaw is just a glimpse of what "next internet" could look like. We need a hackathon. We will also need to rethink SaaS pricing / margins / business in general (Andreas Klinger - good video - thinks only build for enterprise - monetize domain knowledge + trust). Tomas launches Agenica.
• Who is already making money with Agents?
  • Let's define first, it's not just automation. AI agents reason about how to achieve goals: "I need to accomplish Z, let me figure out the steps." They can handle novel situations, adapt when things go wrong, and choose from multiple possible paths.
  • Pactum AI. Ace Waves. Evergrowth. Chaseit. Kopa AI. Nexos.ai. Sera AI. Desktop Commander. Now in the enterprise context, this often requires to fix the foundations first and only then scale AI – that's why Ace Waves applies forward-deployed delivery model.
• Deeptech. Would love to see this across Baltics: TalTech list of technologies that need CEOs to turn them into businesses. Catalyco was pitching at LitBAN (sustainable and local supply chain for chemical industry).
• Health. Martynas in winning mode with Livity. Pitched on Shark Tank in December, said yes to €100k for 40% – but deal broke down in negotiations; in January: Livity hits €98k in one month.
• People. Gražvydas Kaminskas steps into CMO role at Ondato. Roman Lobas joins Icebreaker.Agency as a Co-Founder and Partner. Alius Petraška joins Omnisend as Engineering Director. Mindaugas Mačėnas joins CUBE3 AI as Head of Finance. Andreas is now Head of Commercial at GIFQ.
• Calendar. Hostinger: Code on Tap, Wednesday evening at Sparta.
  Join Future of Work in Riga: how we’ll lead, learn, and stay human in the AI era?? March 16, 2026.
• Emerging scale-ups: firms that hired most in Lithuania in year 2025. Only looking at firms 30-200 people size, and headcount in LT. Surprise surprise – most are actually bootstrapped.

‍

rounds and capital

• Onodrim launches with EUR 40M seed by Founders Fund, Lakestar, and General Catalyst
• Farsight Vision finalized €7.2M seed round - to accelerate AI-Driven Decision-Making in Defence. Investment by SmartCap, Axon Enterprise and European defence investors.
• Tingit closed a €1.5M seed round – led by Coinvest Capital and joined by FIRSTPICK, LitBAN, NGL Ventures, and previous investors Heartfelt_, BADideas.fund, and Purpose Tech. At 14K users, operating in Lithuania, France, Poland. Everything from €20 sneakers to €15K Hermes bags.
• Deep Space Energy secures €350,000 in investment – Outlast Fund + Linas Sargautis (+ grants and contracts, total €930,000).
• Exit: missed this out, so only 8 months late. Zygimantas have co-founded Ro5, proprietary AI drug discovery company, which was acquired by Juvenescence in June 2025. Impressive!

brand + design

w/ rokas from achoo

• Big Taste. Paul Graham joins the chorus: taste is now in demand. It’s an era of infinite output. Infinite output doesn’t mean infinite progress. Selection becomes the work.
• Simple way to improve your design game. Spin-the-chair test your interface design.
• Brand at civilization scale. Joe Gebbia (Airbnb co-founder, now U.S. Chief Design Officer) shares early progress from the National Design Studio redesigning federal services across digital and physical spaces.

founder's guide

• Chipping Away. Very good distinction for early-stage building. It's not making a painting, it's closer to crafting a sculpture.

Under the painting model, a failed experiment is a mess. You put bad paint on the canvas and now you have to figure out how to cover it up.

Under the sculpture model, a failed experiment is progress. You chipped away a piece you didn't need. The sculpture isn't visible yet, but you're closer to it than you were before.

One model treats things that don't work as problems. The other treats them as progress.

further insights

• Something Big Is Happening – this essay travels around now
• Mindaugas, at Lovable, writes – The best education in the world is now $25.
• If you have friends with careers in Customer Support - help them and encourage to change lanes, these jobs are disappearing
• The Ground Is Moving – by Kristjan (Arbonics)
• Show me ridiculous graph on startup growth? Claude in 3 years

‍