Automation Diagnostic
5 business days
What's included:
- Two working sessions with the people who run the process, not only with management
- A map of your operation showing the hours each step consumes
- The 5 most automatable tasks, ranked by hours per month you get back
- A working prototype of opportunity #1, running on your real data
- A 30/60/90-day plan with costs and owners
When you're ready to scale.
These packages are quoted after the diagnostic, scoped to your operation.
Pilot
Process #1 from the diagnostic, in production and running on its own. 30 days. The diagnostic is credited against the price.
AI Operations
Your automations kept alive: monitoring, adjustments when your operation changes, and a new process every quarter.
Fractional AI lead
I run your automation agenda without you hiring a CTO. Priorities, vendors and team.
The diagnostic is always the first step and the only one with a fixed price: when it ends you know exactly what the next step costs, because we have already seen your operation from the inside.
Two jobs almost nobody else takes.
Anyone can automate a process. Opening up a fifteen-year-old .NET system without shutting it down, or standing up a language model on your own server, takes having done it before.
.NET modernization + AI
Your fifteen-year-old system still runs the business and nobody dares touch it. I pull it apart in pieces, without taking the operation down, and add a copilot that answers questions about your own data. WebForms, .NET Framework, legacy SQL Server.
- An inventory of the monolith: what can be split off first, what stays put
- Module-by-module migration, with the old system running alongside
- An internal copilot over your database, not over generic documents
- Your team is trained to carry on without me
CEMEX, Ternium and NEORIS: twenty-six years of systems still in production.
Ask about this workSovereign on-premise AI
Language models running on your own server, with Ollama. No case file, no statement, no contract leaves your network. For firms, clinics and lenders where shipping data to a foreign API is simply not on the table.
- Llama, Mistral or Qwen on your hardware, with no per-token monthly bill
- Search across your internal documents (RAG) under the permissions you already have
- Meets data-residency requirements because the data never leaves the building
- Fixed infrastructure cost, not an invoice that grows with usage
Installed in your data center or your private cloud. You keep the model weights.
Ask about this workHave questions before deciding?
Message me on WhatsApp and I'll reply directly — no bot, no forms.
💬 Ask me on WhatsApp