At our third meetup we opened with who this group is for, ran a live AI trivia round, and heard from two very different builders: one tackling energy infrastructure, the other small-business bookkeeping.
Who we are, and who we're not
We started by setting the tone. Building with AI Charlotte is for builders at every level, whether you're already shipping with agentic tools or just getting started. It's a place to share what you're making, trade what works and get honest feedback. It is not an experts-only club, not a sales pitch, and not exclusive: everyone is welcome, whatever their background or experience, and no question is too basic.
Pitch: Gridramp
Jack Leesnitzer, co-founder and COO · gridramp.ai
One platform to develop renewable and data center projects. Before a solar farm, battery site or data center gets built, there are years of permits, land deals, grid connections and financing, spread across dozens of organizations. Jack spent years on the development side of solar and kept finding the real status of hundred-million-dollar projects living in spreadsheets, shared folders, group chats and someone's inbox. Gridramp puts all of it in one place, so it's trackable and accessible to AI.
The team. Jack, based in Charlotte, brings the developer's view of land, permitting, interconnection and engineering. His co-founder and CEO, Denny Cmiel, in Portland, Maine, was previously VP of Product at Looma, where he helped grow the business from $1M to $10M in recurring revenue, and he architects the platform.
Listen before you build. Before writing code, they mapped how a project really runs from start to finish, then held dozens of conversations with developers. No pitching, just "how do you track this today, and what's painful about it?" The big finding: land, permits, the grid and financing all run at once, and things fall through the cracks in between.
How they build with AI. Every feature starts with a customer need. Next comes a "grill me" step where Claude interviews them one tough question at a time until the idea holds up. They mock up screens in design tools before any code exists, and only then do coding agents in Cursor build it inside a local sandbox. Their summary: "Agents nail the concept. Founders execute." The agents work inside a backend Denny designed, so speed doesn't cost reliability.
AI inside the product. Hand Gridramp a project spreadsheet or data room and ask it to set the project up. It asks clarifying questions first, like a contractor confirming the plan, then builds tasks, budgets and dependencies in a draft with every change tracked. Nothing goes live until a person publishes it. "The AI does the busywork. A human always makes the call."
Go-to-market. Pilots with people who'll give honest feedback, using the product themselves constantly, then launch partners running it on real projects. "There are no shortcuts to a frictionless product."
The ask. Warm introductions to anyone who develops renewable, battery storage or data center projects.
Pitch: ClearlyKept Express
Expert bookkeeping cleanup, without the discovery call. ClearlyKept Express is aimed at growth-minded founders on QuickBooks Online whose books haven't been reconciled in 30 days or more. It's launching in December 2026.
Go-to-market. A product-led funnel built around a free "Books Health Score" diagnostic, partnerships with tax CPAs and M&A brokers, and a QuickBooks App Store listing.
The ask. Three to five Charlotte founders on QuickBooks Online to try the free three-minute Books Health Report beta, introductions to local tax CPAs tired of messy client books at year-end, and coffee chats on product-led growth for small-business services.
Takeaways
- Discovery still beats code. AI makes building fast, which makes building the wrong thing fast too. Talk to users first.
- Make the model argue with you. An AI that interrogates your plan is often more useful than one that writes it.
- Draft, then approve. Let AI do the busywork in a draft and keep a person on the publish button. We're building a whole workshop around that idea.
- Make a specific ask. Both pitches named exactly who they wanted to meet, which makes it easy for the room to help.