Getting an assistant into production
Tool calling, write safety, evals and cost control on .NET
I take on a small number of AI integration projects alongside my main role. Remote, in English or Spanish.
Best suited for teams with an LLM feature that already works in a demo and now has to survive contact with production data.
Is this a fit?
Yes, if...
- Your assistant works in a demo and stalls at the point where it has to write.
- You need tool calling wired to real domain services, with the permissions holding server-side.
- You are running model calls in production and cannot say what they cost or when they regress.
No, if...
- You want the assistant to have unrestricted write access to production data.
- You need a model trained or fine-tuned rather than integrated.
- You want a chatbot bolted onto a FAQ page.
Direct contact
hey@alexmdevs.com
Turnaround
A couple of working days, usually less.