The brief.
Protocol Labs brought me in two weeks before their inaugural pre-accelerator was due to launch. The 20 teams — early-stage builders moving from hackathon prototypes to investable, viable products in the PL ecosystem — were already selected and waiting. What didn't exist was everything they would experience: no curriculum, no engagement framework, no onboarding. And the date was not going to move.
The brief, in practice:
- Two weeks to launch. Prep began the last week of July; the program launched August 13 with curriculum, engagement strategy, and onboarding shipped — on schedule.
- Change as a constant. Halfway through, the program changed — and kept changing. Fully-async became hybrid. One gauntlet became two, then Gauntlet 2 moved up with six days' notice. Demo Day relocated from San Francisco to Dubai. KPIs evolved mid-cohort, and the end date stretched from November 12 to December 10.
- Scale, then focus. Gauntlet 1 in September qualified all 20 teams on traction, product clarity, business model strength, and pitch quality — establishing the top 10 that advanced toward Build Week. Every team was supported the whole way; the strongest earned the Dubai stage.
- Geographic reach. The cohort needed to converge across continents — online program, Buenos Aires residency, and Dubai Build Week.
- Quality bar. KPI targets across workshops, assignments, mentorship, and pitch submissions — every one of them tracked and reported from week one.
The Build Week ten: Treetino (ex-Nocena), Pomodoki, Pulse (ex-TroveFi), MicroCrop, Viatika (ex-X402), SkyFall (ex-SkyHedge), Derby Fish, and Vaultopolis — plus Fildos and Kiyan, who qualified but couldn't continue due to visa restrictions. Meet the cohort on the official Founders Forge site →
The approach.
The core insight was that pre-accelerators fail when they treat founders as cohort members rather than individuals. I designed the program around 1:1 leverage — every workshop, every fireside, every office hour was structured to compound into specific founder outcomes.
Key design decisions:
- KPI framework with SMART goals per founder, reviewed weekly in 1:1s — an average of 11 SMART goals submitted per team over 12 weeks. Founders knew their own scoreboard from week one.
- Buenos Aires residency — an optional 10-day residency around DevConnect for the 6 teams traveling, organized and led outside the original program scope. Daily pitch prep and mentorship, 4 firesides I hosted with Kevin Owocki, Juan Benet, Cameron Dennis, and Mona Teisler, 3 investor wine & cheese nights, and founder coworking. The strongest teams were selected to present at two DevConnect demo days.
- 13 recorded workshops — legal setup, MVP, community-first GTM, token design, smart-contract security, fundraising — plus 15+ IRL pitch-practice sessions. Recordings became evergreen alumni assets.
- Mentorship engine — 60+ mentor sessions beyond office hours, running from onboarding through Build Week, with a bench including Justin Melillo, Masa Kikuchi, Lindsey Winder, Shai Perednik, Ali Serag, Antonis Argyros, Maria Yarotska, Katie Narain, Erik Van Winkle, and Berk Ozer. Every team met with at least one mentor, averaging 3–4 mentors per team, and 100% of mentors paired with at least one team (target: 75%).
- Dubai Build Week — nightly pitch practice sessions and daily 1:1 prep, so every team stepped on stage sharp.
The results.
The program exceeded the majority of its KPIs — and every founder-facing metric landed at or above target.
- 100% of active teams completed 85%+ of workshops (target: 80% completing 75%).
- 87% of teams agreed workshops met their needs (target: 75%).
- 100% mentor satisfaction (target: 75%) — every mentor rated their experience excellent.
- 100% of teams submitted their final pitches on time (target: 95%) — even with Gauntlet 2 moved up on short notice.
- 10 of 11 founders (90.9%) selected office hours with Mel as the most helpful element of the entire program — the #1 result in the exit survey, ahead of Build Week (72.7%) and every other category. Over 100 office hours delivered across 12+ weeks.
- 10 of 11 founders would recommend Founders Forge to another early-stage founder, and 10 of 11 rated the program 4 or 5 out of 5 for overall helpfulness to their startup.
- The top-rated role in the program. 9 of 11 founders rated my support "extremely helpful" in the exit survey — the highest rating given to any role across the entire program.
- 5 teams generated direct investor interest through the Argentina residency — investor relationships built in Argentina, weeks before Dubai.
The highlights worth telling.
The numbers are real but they don't capture the texture. Four moments mattered most:
The two-week build. There was no program in late July. There was one on August 13 — curriculum, engagement strategy, onboarding, KPI tracking. The launch date wasn't moving, so the work did.
The Buenos Aires residency wasn't in the original program. I organized and led it myself — a 10-day optional convergence with daily pitch prep and mentorship, built around firesides I hosted with Kevin Owocki, Juan Benet, Cameron Dennis, and Mona Teisler: honest conversations about how founder ecosystems actually work. The strongest teams earned slots at two DevConnect demo days, and paired with three investor wine & cheese nights, it's where 5 teams found their first real investor traction — before Dubai, not at it.
The gauntlet that moved. The program was designed with one gauntlet; it ended up with two. Gauntlet 2 existed so program leadership could review every pitch before Argentina and confirm Dubai readiness — and it was moved up from late November to November 10 with six days' notice. Six days to announce it and get every pitch ready. I met with ~12 teams daily until the day, on pitch feedback, storytelling, and strategic positioning. Every team showed up ready.
Dubai sharpened everything. Build Week ran on nightly pitch practice and daily 1:1s. By the end, the pitches weren't performances — they were conversations the founders had already rehearsed with real stakes. Every team left prepared and confident.
The engagement closed with a playbook, not just a report. I delivered Protocol Labs a full program report — what worked, what didn't, and concrete recommendations for future cohorts: align KPIs and roles before launch, build the investor pipeline before Gauntlet 1, fix gauntlet timing early, and preserve the high-touch founder support that founders rated highest. Systems that outlast the engagement.
Where the cohort went next.
The strongest evidence arrived after the program — published on Protocol Labs' own Founders Forge site:
- Viatika won $52,500 from Flow and USDFC sponsors, and has achieved rare regulatory breakthroughs in the US & EU — helping shape early legal pathways for x402 payments.
- Treetino earned a $31,000 prize from Protocol Labs and Flow on top of a government-funded prototype, and is deploying test installations across the European Union.
- MicroCrop won $26,500 from Protocol Labs and WeatherXM, and is launching pilot programs bringing transparent, automated crop insurance to underserved farmers.
- DerbyFish is generating revenue from live fishing derbies — proof of demand for on-chain sports tournaments.
That's $110,000+ in prizes and sponsor awards across three teams alone — from hackathon prototypes in August to funded, shipping companies by winter. Watch the cohort recap →
What the exit survey said.
Office hours with Mel — the most helpful part of the program. — selected by 10 of 11 active-team founders (90.9%) in the exit survey, ahead of every other program element
Rated "extremely helpful" by 9 of 11 founders — the highest support rating of any role in the program. — exit survey, "How helpful was support from the following roles?"