Co-Founder @OpenMercato Exited founder of: @DivanteLTD, @callstackio, @useAlokai BA @elevenlabs @medusajs @wordware and 40 more..

Wroclaw, Poland
Joined December 2008
Think Lovable for your ERP & CRM… but with your stack and architecture. Open Mercato Sandboxes are live 🚀 - Close your laptop. Agents keep working. - Any tech stack - Jev, Claude Code & Codex - VS Code, Git, live preview & sharing - Free AI + compute credits to start - Compute billed by the minute youtu.be/0jdlGSlZusI?is=ylGN…
11
8
1
104
567,006
No product idea? Fine. Building bullshit just because everyone else is doing startups? Not fine. In 2008, while others chased VC money, we built software for paying clients. Boring? Sure. But we built revenue, dividends, and independence. Then we built products. On our terms. Investors can buy a story. Customers demand results. AI has made building easier than ever. Has it made startup FOMO even worse?
1
5
885
Tomasz Karwatka retweeted
want to build your GTM engine? apply to The Revenue Hackathon: Berlin by Hubspot, OpenAI, Clay & Lemlist 18 mentors, tokens, tool access - just for 15 cherry-picked early stage startups offers.hubspot.com/berlin-re…
1
1
9
495
I book meetings with myself, then decide that guy can wait. Last 3 months: 33 Deep Work blocks. 16 overlapped with other commitments. Every one made sense. That’s the trap. Starting out, you fight for opportunities. Later, you fight to turn good ones down ;)
1
5
757
Three. That’s how many AI tools a human can juggle at once. Beyond that, productivity drops and serious errors rise by 39%. Now look at your coding setup. How many agent sessions are open? How many have you lost track of? Until now, the orchestration layer was… you. Reading logs. Answering agents’ questions. Merging PRs. Congratulations. You automated coding and gave yourself a full-time babysitting job :) That’s why we built Cezar, our open-source tool for managing coding agents. Try it in our cloud with zero installation, or grab the open-source version on GitHub. cezar.run
4
1
1
4
880
Tomasz Karwatka retweeted
Pojebane czasy dla programistów, jeśli ktoś wdroży u siebie poniższy orkiestrator to może w zasadzie zostawić 1 programistę nadzorującego, bo cezar - przetwarza tickety (ClickUp, Jira etc.) i pisze kod - robi Code Review - pisze testy też e2e - robi deploy github.com/open-mercato/ceza…
27
6
100
27,627
Kto chce pierwszy dostać moją nową książkę do testów? Prowadzisz firmę dłużej niż 5 lat? Firma rośnie, ale Ty nadal zatwierdzasz, poprawiasz, sprzedajesz i odbierasz telefon na urlopie. Brzmi znajomo? +5 - o kolejnych latach rozwoju i budowaniu firmy samobieżnej. Takiej, w której ludzie potrafią podejmować dobre decyzje bez czekania, aż founder wróci z urlopu. Chcę sprawdzić te rozdziały z ludźmi, którzy mierzą się z tymi problemami. Dlatego wybiorę 55 osób, które przeczytają pierwsze, próbne wersje i powiedzą mi wprost: czy to ma sens :) Testerów wybiorę spośród osób zapisanych na waiting listę nowego wydania The 5. cart.easy.tools/checkout/bra…
6
13
1,834
“He must have had connections.” Damn right. I spent 10 years building things online for free. People noticed. Open source. Side projects. Showing up. Doing the work. I bootstrapped my network, too. Don’t confuse earned trust with inherited privilege.
1
2
21
1,115
Coding agents that work 24/7. On your own server. Queue the backlog, close the laptop, review the PRs from your phone. cezar.run
7
2
68
177,399
Nobody who tries our agent orchestrator asks how many agents it runs. They ask how many hours they have to sit next to it. So we scored 11 tools (Claude Code, Codex, Copilot, Antigravity, Cursor...) on one thing: how much of your attention they give back. Full breakdown, ranking, and the gap list we keep public for Cezar. openmercato.com/blog-posts/y…
7
16
166,881
Piszę pięć nowych rozdziałów do The 5. +5 czyli kolejnych latach rozwoju i budowania firmy samobieżnej. Takiej, która rośnie też, gdy jesteś na wakacjach. 1) Jak zadbać o siebie jako founder. Żeby rozwój firmy nie odbywał się stale kosztem Ciebie. 2) Jak budować nowe produkty. Silnik wzrostu z pierwszych 5 lat w końcu staje się wąskim gardłem. 3) Jak zwiększać wartość firmy. Niezależnie od tego, czy chcesz ją sprzedać, czy prowadzić przez następne dwadzieścia lat. 4) Jak mądrze wykorzystać AI ale też być odpornym na AI. Tu praca wspólna bo wszyscy się tego uczymy. 5) Jak przekazać władzę w firmie i co robić po :) cart.easy.tools/checkout/bra…
3
26
60,719
Tomasz Karwatka retweeted
if there was only a safe, clean, modern city with great food, nice people, good airport, builders and ecosyst… oh wait
52
22
3
724
125,898
Weekend inbox check. Most emails awaiting my reply are from…. my own AI agents. I’ve automated giving myself more work. Great job Tomek! “Human in the loop” = me doing cleanup on Sunday. Anyone getting actual results AND less work? Show me what you stopped having to do.
23
42
73,276
I’m probably richer than a 16-year-old Taylor Swift fan. She’s still a better customer than I am. I’ll play a few songs on Spotify. She might spend 5k for a concert. I used to make this exact mistake in sales: confusing a client’s wealth with their willingness to pay us. A big brand showed up and I mentally spent their budget before we’d even sent a proposal. “Wow. Coca-Cola!” Smaller companies often ended up spending more with us. Our work was closer to how they made money. Closer to the problem keeping the owner awake. At a large company, we could be one of hundreds of vendors. At a smaller one, we could help fix something holding back the entire business. That changes the budget. And how long the relationship lasts. How much does solving this problem matter to this customer? A famous logo looks great in your pitch deck. Being essential gets you paid.
3
16
1,593
Looking for 5 new global implementation partners for Open Mercato before the end of the year. Q4 for Open Mercato means one thing: going global. Open Mercato is an AI-supportive CRM / ERP framework. It ships with 80% ready. Someone has to build the last 20% for the client. That someone is a partner. The model is simple. We sell the enterprise license. Partners do the implementation and keep the services revenue. We certify, train and send leads. Oct 15 - Warsaw - Tribes Day Oct 27 - Katowice - European Congress of SMEs Oct 28 - Wrocław - Made in Wro Oct 29-31 - Goa, India - React India Nov 7 - Poland - Forest Planting for Entrepreneurs (yay!) Nov 9-12 - New York - ElevenLabs Summit Nov 15-16 - New York - JS Nation Nov 15-19 - San Francisco Nov 19-22 - Las Vegas Dec 8-11 - Warsaw If you are in one of these three groups, let’s meet.
2
2
27
162,640
Invite your friends to read my Substack. Get AI-Proof Agency Workshop Handout and a Full Mentoring Session with me. tomaszkarwatka.substack.com/…
7
896
The worst business advice I got turned out to be correct. Just not for my stage. Most business advice comes in pairs that cancel each other out. Which common business advice didn't work for you?
2
1
20
2,081
I built and sold two agencies. If I started again, here’s how I’d get the first 10 clients. I co-founded two software agencies. We grew them to 350 and 200 people and sold them in 2021 and 2025. Here’s how I’d do it. CLIENT #1: START WITH TRUST List 10–20 people who know your work. Former bosses, colleagues, people you’ve worked with. Ask what’s broken in their business. Listen before pitching. Offer a small, paid project: - One clear outcome. - Fixed scope and price. - 2–4 weeks. Be honest about your experience. You can be new to running a company and still be good at the work. Charge less if you need to. Avoid free. You need to learn whether someone will pay. CLIENTS 2–10: SELL IT YOURSELF You need to hear the objections firsthand. That’s how you learn what clients actually buy. - Pick the industry or technology where your first projects went best. - Turn results into case studies. Use real numbers. -Ask happy clients for two introductions while the result is fresh. - Partner with a software vendor whose customers need implementation. - Start turning down work that pulls you away from your niche. At Divante, focusing on ecommerce gave buyers a much clearer reason to call us. Another way in: contribute early to a technology with growing demand. Becoming a core contributor gives clients a concrete reason to choose you. Disclosure: I’m now building Open Mercato, an open-source ERP/CRM framework, so I see this from the software vendor’s side too. Also, watch client concentration. One client paying 50% of your revenue has a lot of control over your business. BEYOND 10: MAKE SALES REPEATABLE Ten clients isn’t a magic threshold. But by now, you should see patterns worth documenting. - Write down qualification questions, pricing rules and how you build proposals. - Hire a salesperson once you have something that works for them to repeat. - Publish useful reports, benchmarks and project breakdowns. - Build reusable IP from problems you keep solving. - Grow existing accounts where you can deliver more value. Our work on an open-source ecommerce frontend helped bring international enterprise leads. It later became a separate product company. That took time. Start by making each project leave something useful behind for the next one. A practical test: take two weeks off. Do leads still come in? Can your team move deals forward? Where are you stuck: landing client #1, reaching ten, or making every sale depend less on you?
9
1
27
8,577