Let’s take it easy. Before we meet, send me the problem. I’ll read it and reply with a short preliminary assessment, so our first conversation starts from something concrete rather than “so, tell me about your project.”
Bring a rough idea, a contractor proposal, or a problem you have struggled to frame. If I am not the person you need, I will tell you.
What you’ll get back
- My reading of the problem, and what I suspect sits underneath it.
- Whether AI or ML is plausibly the right tool. Sometimes it is not.
- What I would investigate first, and what is still missing.
- Whether I think I can help, and a suggested next step.
This is a preliminary read, not a solution or a design. I review every submission personally; if it looks like a reasonable fit, you’ll hear back with the assessment and we can decide whether a meeting makes sense.
Request a review
Ten to fifteen minutes of writing. Plain language is better than polish. Answer what you can:
- What are you trying to achieve? The outcome or decision you want to improve, not the technology.
- What happens today? How it is done now, and what isn't working.
- What makes it hard? Data quality, scale, accuracy, latency, cost, integration, human workflow — or not sure yet.
- What have you already tried? Optional.
- Where are you with it? Exploring an idea, reviewing a proposal, ready to build, or already built and struggling.
- Your industry and organisation size.
- What data and systems are involved? Optional. Rough volume, where it lives, anything already running. Guesses are fine.
- Is something driving the timing? Optional. A budget decision, a contract, a launch.
Attach anything useful: diagrams, requirements, a proposal, a small data sample.
Send for review Opens an email with these questions. Kept private, never published.
Prefer to write it your own way? h.sh.farahani1@gmail.com