Grouped by discipline rather than by logo, the stack I actually reach for on each kind of problem.
Web apps with the model built in: API, database and interface shipped together.
LLM workflows that call tools, check their own output, and hand off to a human when unsure.
Training and tuning models for vision, language and prediction, then making them fast enough to serve.
Containers, CI/CD and deployment, so the model runs the same in production as on a laptop.
Send the brief. A 15-minute intro call is free, and I'll tell you honestly whether I'm the right fit, including when I'm not.