Digital Thought Labs
Where AI prototypes become production.
Digital Thought Labs is a research-and-engineering lab that builds AI-first systems and documents the process in the open. Its focus is the unglamorous gap most AI projects fall into: getting from a promising prototype to something that survives production. The tagline captures it — “where prototypes become production.”
What it works on
The lab’s work clusters around the hard middle of applied AI, the part that lives past the demo:
- AI-first system design — architecting products where the model is the core, not a feature bolted on the side.
- Multi-agent orchestration and prompt engineering — getting reliable behaviour out of systems that are inherently probabilistic.
- Production deployment — the monitoring, evaluation, and guardrails that separate a viral demo from a system you can trust with real users.
- Benchmarking — including BuildBench, an effort to measure how AI coding agents actually perform rather than how they’re marketed.
Research in the open
What sets the lab apart is that it publishes. Instead of hype, it puts out original write-ups on AI engineering — case studies from real client projects, technical research, and honest accounts of what worked, what failed, and why. That makes the material useful even if you never hire them: the lessons stand on their own, and there’s a YouTube channel and active GitHub, X, and LinkedIn presence for people who want to follow along rather than commission work.
This “show your work” posture is rarer than it sounds in AI, a field where most public content is either a marketing deck or a breathless demo. A lab willing to write down the failure modes — the prompt that worked in testing and fell apart in production, the agent loop that quietly ran up a bill — is signalling that it actually operates at that level, not just talks about it. For a reader, that honesty is the value; for a prospective client, it’s the due diligence done in advance.
Who’s behind it
It’s founded and led by Dr. Zac Smith, who holds a PhD in Computer Science, has been writing code since 1983, and currently works as an AI Engineer at Gauntlet AI. Past work referenced by the lab spans Gauntlet AI, Best Buy, IMAX, CVS Pharmacy, and dozens of Fortune 500 companies — the kind of track record that lends weight to the “we’ve actually shipped this” framing.
Who it’s for
Technical teams and organisations trying to move AI from experiment to product, plus individual engineers learning the craft. If your prototype demos beautifully but keeps falling over the moment real traffic or real edge cases arrive, the lab’s whole reason for existing is that specific pain.
Why it’s here
Digital Thought Labs shares DNA — and a founder — with the team behind FindMeeting and Code Shock. The same instinct that makes FindMeeting a no-nonsense, no-signup tool runs through the lab’s work: build the real thing, measure it honestly, and write down what you learned. If you’re trying to ship real AI features rather than demos, it’s a good place to read and to talk to — and the guides and free FindMeeting tool are part of the same network.
New here? FindMeeting is the free group scheduler from the same team — or read the guides and see the rest of the network.