Fractional / advisory
Fractional ML engineer and AI advisory for founders and CTOs
Plenty of teams do not need a full-time senior ML hire. They need senior judgment available at the right moments: before an architecture is locked in, before a build-vs-buy decision, before an expensive optimization that will not pay off. That is what an advisory engagement is for.
Who this is for
- ▹You have engineers who can build, but nobody senior in AI to check the direction
- ▹You are about to make an expensive infrastructure or vendor decision
- ▹You are hiring for AI roles and cannot evaluate candidates confidently
- ▹You need an honest read on whether AI is even the right tool for the problem
What the work covers
Architecture and code review
Regular review of the AI parts of your system, with specific, prioritized changes rather than general advice.
Build-vs-buy guidance
Cost and risk modelling with real numbers from your codebase, so the decision is defensible to your board.
Hiring support
Role definition, technical screening and interview design for ML and AI engineering hires.
On-call technical direction
A senior person to think with when something breaks or a decision is blocked.
Evidence this works
- ✓Currently the sole ML engineer building a full production stack for a safety-critical healthcare platform
- ✓30+ AI and ML systems shipped across healthcare, enterprise automation, legal and computer vision
- ✓Published peer-reviewed research in Computers in Biology and Medicine (Elsevier, impact factor 7.7)
How the engagement runs
- 1Intro conversation to understand where decisions are actually getting stuck.
- 2A monthly retainer with an agreed cadence of review sessions and async availability.
- 3Clear written recommendations after each session, so decisions are documented, not verbal.
Common questions
How is this different from a full build engagement?
Advisory is judgment and direction, not delivery. Your team builds, I make sure the direction, architecture and trade-offs are right. If a project needs hands-on delivery, that is a separate build engagement.
Will you tell us not to use AI?
Yes, when that is the honest answer. Sometimes the right solution is a simpler deterministic system. I would rather keep a reputation than sell a project that will not work.
What time zones do you cover?
I am remote-first from Lahore, Pakistan (UTC+5), and work with teams across the US and EU with solid working-hour overlap.
Related work and demos
Working on something like this?
Tell me what you are building and where it is stuck. I reply within one business day.
Send a project inquiry