KWBuilds

KWBuilds Weekly

June 14, 2026

KW tech, week of June 8 to 14

Two founder webinars, DataBraid's pre-seed, and a full events week

This week was a how-to. I spent two afternoons in founder webinars, one on spinning up an AI company while you still have a day job, the other a VC walking through exactly how he decides who gets a cheque. There was a fresh pre-seed round to dig into, a community Claude meetup in town, and a calendar full enough that you could have been out at something most nights. Here's what happened in KW this past week.

DataBraid's pre-seed

A funding story that hit our feed this week is DataBraid, which raised $1.9M USD (about $2.7M CAD) in pre-seed money, its first round, closed in April. It's led by Koru Ventures

, a venture studio backed by the Ontario Teachers' Pension Plan. DataBraid builds software for insurance brokers: instead of logging into a dozen separate carrier portals and re-keying the same data into each one, brokers get a single workspace and the system handles the portal connections behind it. The founders are Nick Romano (CEO, who previously ran Montreal AI startup Deeplite before it sold to STMicroelectronics) and Atif Khan (CTO), and their pilot customer is Scoop Insurance, a brokerage in Welland running around 300 brokers.

DataBraid is a Toronto company, associated with Communitech according to our kwbuilds directory

. Coverage is from BetaKit
.

Two founder sessions I sat in on

The substance of my week was a pair of free Founder Institute webinars, both surfaced on the Waterloo calendar by the local Founder Institute chapter, Waterloo Startup Founder 101. One was about building the thing, the other about getting someone to fund it.

Building an AI company around your day job (Adeo Ressi)

Monday afternoon I caught "How to Launch an AI-Powered Company With a Day Job"

, run by Adeo Ressi, who chairs Founder Institute and runs Decile Group. By his own telling he has founded 11 startups, launched 14 venture funds, had nearly $2B in exits before he turned 30, and "helped create a dozen or so unicorns." He's coming out of retirement to run an in person AI chapter in Silicon Valley this summer, and said he wants to recruit 50 founders and build a few unicorns with them.

There were no slides. He ran the whole session live inside Claude and Replit, and said the best model out there right now is Claude Opus 4.8. KWBuilds runs on Opus 4.8 so I feel seen.

His walkthrough covered:

  • Validating an idea by asking Claude to role-play ten skeptical customers and score it. An audience-submitted idea averaged 3.4 out of 10. He framed this as a stand-in for the customer interviews founders usually skip, done in minutes instead of weeks.

  • Keeping a side project legally separate from your employer: your own computer, your own AI accounts, your own API keys, or the employer may have a claim on what you build.

  • Aiming at what he called "adversarial tasks, not lazy ones," skipping reminders and to-do lists for things like disputing a medical bill, fighting a parking ticket, or chasing a refund, where there's money on the other side and a cut to take from what's recovered.

  • His view on defensibility: the moat is the data and playbooks a company accumulates over time rather than the model, and since standing software up is cheap now, niche markets are worth chasing. His example was a thousand corner stores paying $100 a month.

The live demo never actually rendered, and the whole room watched it happen. He pushed the AI to promise "$1M ARR in 45 days," kicked the build over to Replit in Turbo mode, and it sat on a black screen asking for a QuickBooks API key for the rest of the session. (Is this why I can't afford new pc parts? 😭) The idea validation and reframing he walked through are genuinely worth stealing, even if the "build a few unicorns this summer" pitch around them is mostly there to fill his Silicon Valley accelerator.

How a VC actually decides (Maayan Teper, Resilience VC)

Tuesday I sat in on a second Founder Institute session, a live investor Q&A called "What It Takes to Get Funded"

, hosted by FI's Chris Foltz with Maayan Teper, an associate at Resilience VC. Resilience is a Washington DC seed stage fintech fund, founded in 2023 by Tahira Dosani and Vikas Raj, that closed a $56M debut fund
 last year with LPs including MetLife and the Skoll Foundation. The thesis is fintech that improves financial resilience for lower and moderate income Americans and small businesses, which in practice means they fund US-domiciled companies with US customers. Useful to know before you pitch them.

What Teper got into:

  • A "no" is usually just "not applicable." Teper's point was that most rejections mean you didn't fit the thesis, the geography, or the stage, not that the company is bad. His fix is unglamorous: spend ten minutes on a fund's site and read what they actually invest in before you email. A generic AI-written cold email visibly lowers his interest.

  • The screening is done by associates, not partners. Resilience sees thousands of companies a year, takes a few hundred meetings, and runs real diligence on maybe 30 to 40. The analyst and associate layer is where you get through or don't, so don't write off the junior person reading your deck.

  • A company with no AI in it can absolutely still get funded, and he can tell when founders bolt AI on out of FOMO. Vibe coded products aren't disqualifying either, he just starts asking the normal risk questions about security, scale, and bugs.

  • What he underwrites is the return. Service business or SaaS, he doesn't much care which, as long as it can return the fund, roughly 15 to 20x. His closing advice was that speed and velocity matter more than ever, and to "bring yourself" to the pitch.

Between them the two sessions covered the two halves of this: building the thing, then getting it funded.

Events recap

Plenty on the calendar this past week:

Plus the two Founder Institute webinars I covered up top.

Looking ahead this week

A week with a lot to actually do and learn. See you next week, or at one of the events. Come say hi if you spot me.

Torrin