Cofounder is a 30-day build for founders who already tried and stalled. You bring the idea or the business that has not moved. You leave with a venture running under its own power, an AI cofounder doing the grind, and the customer conversations and judgment that no machine can do for you still firmly in your hands.
You did the program. You pitched. The door closed anyway.
Maybe you went through a free cohort and came out with a tidy idea and no traction. Maybe you applied to Spark or a grant and got passed over. Maybe the business is real but it has been sitting in the same spot for a year because there was never a team, a budget, or anyone in the room who had actually done it.
The advice was fine. The problem was never the advice. It was that you were doing all of it alone, and free programs are built to inform you, not to finish you.
A year ago you really did need the grant, the cohort, the team. You do not anymore.
Most of what an accelerator was gatekeeping is now available to one motivated person at a kitchen table, if they know how to work with an AI cofounder instead of just typing prompts at one.
That is the whole shift. Everyone has AI now. Not everyone thinks with it. Cofounder is the 30 days where you learn to, on your own business, with someone who has built and taught this for years sitting beside you.
Plenty of programs will help you polish a plan. This one is built to put a venture in motion in 30 days and hand you the system that keeps it moving after.
You finish with conversations you actually had and an experiment you actually ran, not a deck full of assumptions.
Research, modelling, copy, landing pages, decks. The work that used to need a team, compressed into days and directed by you.
You leave with the templated cofounder workflows themselves, not just four weeks of lessons. That is the part competitors cannot copy.
An async lesson lands before each live session, so the live hour is pure work on your business. One on-ramp, then four weeks. Every stage names what the cofounder does and the one place you cannot trust it.
Talk to real people until you know the pain is real. New founders run this in full; if you already have a business, you audit it fast.
Drafts the market scan, the interview guide, and the personas, then synthesizes your real transcripts into cited themes.
It cannot invent customers. AI-generated "validation" is the one move that voids the whole thing. The proof comes from real voices only.
Own the judgment. Decide which one or two assumptions the whole business rests on, and hold the skeptic's seat.
Runs the rigour that used to be slow: market sizing, unit economics, value proposition, the assumption map. MIT-grade work at lean speed.
Every number it gives you is a hypothesis wearing a suit. You mark the two that matter as claims to check. Verify the number.
Send the outreach as yourself. Have the conversations. Run one real experiment against your riskiest assumption. This is why it is 30 days and not a weekend.
Builds the landing page, the outreach, the survey, the experiment assets, and stands up a working page from a prompt.
It cannot be in the room. The customer conversations are yours, full stop. Everything else flexes; this does not.
Supply the real numbers and the real traction. Stand up and tell the story. The pitch is yours; the cofounder is the editor.
Turns your value proposition into a one-liner and a full deck, generates variations, tightens the language.
It will fabricate financials, traction, and team slides if you let it. You run a line-by-line pass so every figure traces to something real.
Read your real signal for what it is. Decide what counts as traction. Keep your hand on the wheel of every workflow you retain.
Pulls the signal into a 90-day roadmap and codifies the workflows you keep running after the program ends.
"Set and forget" agents are oversold. Your keeper workflows are documented as a supervised partner, never an autonomous employee.
The honest version of an AI program tells you where the tool lies. The cofounder does the production. You do the thinking that cannot be offloaded. Three rules run through every week.
Research, models, copy, assets, decks. The compression is real, and it is where AI earns its place in a one-person company.
No synthetic respondent stands in for hearing a real customer hesitate. The program lives or dies on refusing that shortcut.
AI output reads as finished and authoritative. You leave with the habit of labelling what is proven and what is still a guess.
I won this myself. About ten years ago I was the founder in the arena, not only the person at the front of the room.
I have coached more than one Spark winner. Producing winners repeatably, not once by luck.
I ran the two-week, multi-institution bootcamp format for roughly 100 students across Nova Scotia post-secondaries at Acadia, before national programs started selling it as new.
Seven years teaching at CBU, thousands of founders through Mashup Lab and CBU. A decade-long track record, now rebuilt with AI.
A founder from my AI-integrated workshop won CA$50,000 from Spark NS. The live, AI-era proof point.
I helped build the build-and-go-to-market arc that several free programs still run. Insider authority over the thing the market gives away.
No. New founders run the full on-ramp and arrive at Week 1 with a validated problem. If you already have a business that stalled, you move through the on-ramp fast and spend your energy getting it moving again.
One live working session a week for four weeks, plus an async lesson before each, plus the real work on your own business, most of it in Week 2 when you are talking to customers. Plan for real effort, not a passive course.
No. The whole point is that the AI cofounder does the building. You direct it. You bring the judgment and the customer conversations.
No. Tool names change every few months and are kept to a thin appendix. What you keep is a method and a set of workflows, plus the discipline to know when the tool is lying to you.
A business in motion with real signal behind it, a 90-day plan built on what is proven rather than guessed, and the templated cofounder workflows you keep running long after the 30 days end.
Because I run every session myself and the founding cohort gets real attention per person. It also keeps the first round honest, so the testimonials mean something.
Tell me in a sentence or two what you are building or trying to restart. Founding seats are first come, first served, and there are only five.
Applying does not lock you in. We talk first to make sure it is a fit.