Sinan Onur Altınuç, PhD · AI services

AI works in the demo. What about in your business?

The way to find out is to try it on your own documents and records. I've spent 16 years building AI systems in industries such as banking, finance and defence. The first call is free; if AI isn't the right answer for your problem, I'll tell you then. If it is, we start with a small project, and you decide on anything bigger once you've seen the result.

  • 16+Years of Experience
  • 20+Completed AI Projects
  • PhDCognitive Science
  • Former DirectorAI & R&D
Prometheus AI

About Me

I've shipped enough AI systems to know the model is rarely the hard part. The hard part is everything around it: the inputs nobody planned for, the checks, the day a model update changes the output. That's the part I build, alongside the models.

I have spent 16+ years in AI and R&D, including as Director of AI & R&D, and I hold a PhD in Cognitive Science. I have built systems in computer vision, natural language processing and data science for banking, finance, defence, aviation and space. Prometheus AI is one person: you talk to me, and I do the work.

If any of these sound familiar, I can help

16+ years of building AI systems, and one person you talk to directly.

  • Staff spend hours reading, checking or re-typing documents

    I build systems that read and enter them automatically, with a person reviewing only the cases the system is unsure about.

  • A task with cameras or images still needs a person

    Inspection, counting, sorting and monitoring can often be automated. I test it on your own images first.

  • You have data and a question, but no one to model it

    Demand, forecasting and planning questions answered with models, delivered as plain reports or dashboards.

  • An AI pilot works in the demo but not on real data

    I measure how often it is wrong on your own cases, find out why, and fix what is worth fixing.

  • Your team checks every AI output by hand

    I add checks that show when an output can be trusted, so people review only the rest.

  • You are choosing between AI vendors or models

    An independent test on your own data, side by side, before you buy.

How We Work

We start with a free call, and at every step you decide whether to go on.

  1. Free Intro Call

    Thirty minutes on your system, your data and the decision in front of you. If AI isn't the cheapest reliable way to solve the problem, we tell you on that call.

  2. Scope and Fee

    If we go on, we define a small first project. Its scope and fixed price go into a one-page proposal before any work starts.

  3. A Test on Your Own Data

    Usually one to three weeks. You provide a contact person and sample data. You get where the current process fails, a test set from your cases with a measured baseline, and success criteria in writing.

  4. A Decision Based on the Result

    We go through the result together: go or no-go, with a rough cost for the next stage. Whether to continue is your decision.

  5. Build and Hand-over

    We build in fixed-price stages, each accepted on measured results against the agreed bar. The code, the documentation and the test set are yours.

Book a Free Intro Call

The call commits you to nothing. If we go on, we agree the scope and fee before any work starts.

Or reach me directly