Agilitee

I rebuilt the entire power plant in my '63 Chevy Impala using AI. I'm not a mechanic. I didn't watch one YouTube video. I stood over the engine bay, took pictures, and had Claude walk me through it step by step. 200 hours of work. It felt like playing with Legos again. Then AI recommended a battery isolator switch - which I installed. Weeks later, I asked it where to connect a new piece of electrical equipment. It came back: "Wait, why did you install that isolator switch?" "Because you told me to." Now there's a 1963 Chevy Impala driving around New Hampshire with a battery isolator switch targeted for a nuclear submarine. AI is confident even when it's wrong. Always have humans in the loop. #AI #AITransparency #TechLessons #ArtificialIntelligence

6 engineers. 6 months. $1M. Then one engineer. 6 days.

Thank you, Michelline Dufort, for bringing us together for a practical conversation about AI in business! At Agilitee, Fortune 1000 companies, including retailers, hire us to use AI to solve real business problems. Better customer experiences. More effective teams. Measurable results. And sensible guardrails around data, risk, and accountability. Family businesses deserve that same practical approach. Start with a problem worth solving, give people the support they need, and measure whether it’s working. Christine List, thank you for sharing your expertise and the stage with me. And Sullivan Street - my son and colleague at Agilitee - thank you for the help throughout the presentation. Doing this with you means a lot son! And thanks to Dan Yarrington for hosting us at Tycoon Games! (What a cool business) I’d love to bring this conversation to more CIO, CPO, and leadership events. There’s a lot to talk about!

One of his engineers barely commits code. In a single day, that engineer burned 30 days of token spend. He pushed back. The partner said the output justified it. This is a VP of technology at a Fortune 30 company. It's the offshore shop building his software. So he pays twice. Once for the bodies. Again for the tokens those bodies burn. And the work keeps getting worse. Everybody blames the AI. The AI's fine. Nobody's driving it. Turn an agent loose on a big codebase with no guardrails and it spins. It spits out work somebody has to redo. That's your bill. The spinning, not the shipping. Work with us and you get Ckye. It feeds the AI only the context a job needs and flags bad work before it ships. You don't pay extra for it, and we don't mark up your tokens. We get paid when your software gets built. 84% of our code gets accepted on the first try. Do that, and the token bill sorts itself out. If your AI spend is climbing while your output drops, ask whoever builds your software one question. Are you charging me for hours, tokens, or results?

"We made our developers faster with AI." Cool. Now your compliance team is buried. The fix isn't more reviewers. It's making compliance a property of the output, not a step at the end. Encode the standards. Check every output against them automatically. Flag drift before it ever reaches a human gate. Audit trail the whole way. Now compliance signs off on the 85% that's already clean and spends its time on the 15% that actually needs a human. Stop treating compliance like a net at the bottom of the cliff. Build the guardrail at the top. Is your compliance team reviewing AI output, or cleaning up after it?

“Just port it over.” Technology partner sees an opening. New need, new deal, new revenue. Everybody at the table wants speed. So someone says it: just port the old app to the new platform. You know what you get? A win on the sale and a loss on every other metric that matters. Conversion. Key KPI. Customer satisfaction. Retailer ROI. The hardware shines. The experience embarrasses it. Here’s what changed in the last 18 months: defining, designing, and shipping a modern experience used to take 9 months. Now it takes weeks. AI-assisted design and code collapsed the timeline that “just port it” was solving for. The shortcut isn’t a shortcut anymore. It’s just shortsighted. If you’re upgrading the tech, upgrade the experience. More importantly upgrade the thinking.