Built from real problems. The ones I had at 17, 19, and 3am last Tuesday.

Everyone has opinions about football. I wanted receipts. So I built Top Bins, a prediction engine trained on 49,000 international matches, running an Elo rating system into a Dixon-Coles model to actually price in scorelines, not just vibes.

Then I called all 104 games of World Cup 2026 before a ball was kicked, and posted every single one publicly so there was no hiding.

Here's how the knockouts landed:

• → Quarter-finals: 4/4 correct. Every single one. No hedging, no "well I said it could go either way."
• → Semi-finals: 2/2 correct. Both calls, on record, before kickoff.
• → Final: called it. The one everyone actually watches. Got it.
• → 250,000 views on Instagram while it was happening, because predicting in public only counts if people can watch you be wrong.

The whole point was never to be a guy with hot takes. It was to see if the math could out-predict the noise. Turns out, trained on enough history and priced properly, it can.

Built for the ones who'd rather be right on record than loud in the group chat.

I learned this the hard way at 17, sending awful, generic emails to people I admired, getting ignored, then giving up. Years later at Roots Funding, I watched founders raise capital off a single well-crafted email to the right investor. Same person. Same hour in the day. Wildly different outcome.

The difference wasn't volume. It was that they sent emails that actually felt personal.

So I'm building 20 Conversations a Day, an AI tool that writes hyper-personalized cold emails using:

• → Proven templates from successful outreach to people like Elon, Bezos, Patrick Collison, and YC founders
• → Real-time context from each recipient's recent podcasts, articles, X posts, and IG content
• → Your own background, so the email actually sounds like you

The goal isn't to send more emails. It's to send the 20 emails a day that should've been sent, the cold reachouts to the dream investor, the founder you look up to, the recruiter at your dream company, the journalist you want to write about you.

Currently in beta. If you want in, send me a cold email. Bonus points if you used the tool to write it.

My first year at Sauder, I had four jobs, three group projects, two clubs, and zero idea what was actually happening before 9am. Slack pings, Canvas deadlines, KPMG emails, interview invites, all of it scattered across six different apps, and I kept missing the ones that mattered.

So I'm building YourFirstHour, an AI morning brief that pulls from your email, calendar, Slack, and Canvas and tells you what actually matters in the first hour of your day.

What needs a reply today. What deadline is closing. Who you said you'd follow up with. What meeting you should actually prep for.

It's not another productivity dashboard. It's the first 60 seconds of your day, written for you, so you can stop opening seven tabs at 8:47am and start your morning knowing what to do.

Built for students drowning in inputs. Quietly hoping working professionals find it too.

Every founder, builder, and student I know has the same problem, they're doing interesting things daily, but the gap between "had a great call" and "posted about it on LinkedIn" is somehow infinite. The wins go undocumented. The lessons get forgotten. The audience never grows.

I tried writing posts manually, and they took 30 minutes each. I tried AI tools, and they all sounded like LinkedIn corporate cringe with three emojis and a "thoughts?" at the end.

So I'm building LinkedInnit, an AI tool that takes your raw thoughts, voice notes, or quick wins, and turns them into LinkedIn posts that actually sound like you.

Not like ChatGPT. Not like a thought-leadership farm. Like you, after one coffee, telling your friend a story.

Built for the people who are doing the work but losing the narrative.

I co-founded Cents in 2023, an advertising-led model where consumer goods like water bottles became billboards for brands, and customers got them for free in exchange.

The idea sounds simple. The work was not.

We raised $5,000 in pre-seed, filed provisional patents across Canada, restructured the cost stack down by 30%, and hit peak month-over-month growth of 25%. I learned how to design an entire operating system from zero, vendor sourcing, payment processing, brand partnerships, unit economics, at 19, with no playbook.

Cents wound down in 2025, but it's where I learned everything I now use:

• → Brand placements only work when the product is something people actually want
• → The hardest part of a startup isn't the idea, it's the unsexy infrastructure underneath
• → Distribution > product, every single time
• → Patents are useful for credibility, less useful for moats

The advisory side of Cents is still alive, I now help SMBs design AI-enabled workflows for the same parts of the business I had to figure out by hand.

my first ever pitch deck →