Strategy
Research & Insight
Find the real constraint. Map the process, question the assumptions, and turn what you learn into a clear first move.
Make AI workflows useful. Make data trustworthy. Explore the work, learn the method, or bring Kenny a problem to solve.
Three reproducible demonstrations using synthetic data. Inspect the method, download the files, and explore the limits. These are not client case studies.
Run the example, interrupt it, and inspect how it recovers.
Automation · Code + recovery logDownload the messy dataset and reproduce the cleaning decisions.
Data · Dataset + audit trailInspect answer checks, failure cases, and the limits of the evaluation.
AI · Test set + evaluationAbout DWKD
Start with a real question. Build something useful. Then show the work: what it does, what you learned, and how someone else can try it. That’s the method behind DWKD.
Learn the method and build something of your own. Or bring me a complicated problem and we’ll work through it together. The starting point is different. The curiosity is the same.
A useful idea deserves a working version. Turn what you discover into a tool, a workflow, or a clear decision. Show what works, be honest about its limits, and use the result to choose what comes next.
Our services