Robiul Islam

Founder, JobForesight

I built JobForesight to answer one question: whose role AI actually reshapes, and what they should do next — the honest read I'd give a close friend, not a sales deck.

I've spent twenty years working in product and data — close enough to AI systems to know what current models can really do, and close enough to working professionals to know how scary the headlines feel when nobody can tell you what they actually mean for your Tuesday morning.

Most "future of work" content sits at one of two extremes: doom for clicks, or vendor optimism that conveniently sells you a course. JobForesight tries to live in the middle — specific time windows in months not years, occupation-by-occupation rather than sector hand-waves, and an explicit list of what the model does not claim. If you want to see exactly how the numbers are produced, the methodology page walks through every input.

Why I built JobForesight

The conversation about AI and jobs has a credibility problem. Consultancies publish glossy reports built for board decks. Tech companies publish capability demos that double as marketing. Newspapers publish whichever framing wins the click that week. Almost nobody writes for the actual person trying to decide whether to invest two years in becoming a better paralegal, a better accountant, or a better account manager — and what the alternative looks like if the answer is "probably not."

I wanted a tool that respected that decision. So JobForesight scores AI exposure across 333 occupations, expresses time horizons in months rather than years (year-level precision is suspiciously round), and grounds every score in published research — O*NET tasks, the Anthropic Economic Index, ILO and BLS data, Microsoft Copilot usage studies — combined with editorial judgement that's spelled out, not hidden.

The core promise is small but real: when you read a JobForesight page, you should come away with a clearer view of which parts of your role are exposed, on what timeline, and where to put your effort next. Not certainty — nobody honest can offer that — but a structured opinion you can argue with. The full scoring model lives on the methodology page; the rest of the site is what that model produces.

Career background

Twenty years in product and data, with a working career split between hands-on engineering and senior product leadership.
Senior product roles in data-platform and analytics companies — building tools that quantify things which feel un-quantifiable.
Consulting on data and AI strategy for enterprise teams making real headcount and capability decisions.
Founder of Solid Computing Ltd, a UK company that builds opinionated software in the AI-and-work space (JobForesight is its current product).
Long-time student of the labour-economics literature on automation — Frey & Osborne, Acemoglu, Autor, and the more recent task-level work coming out of Anthropic and OpenAI's economic research teams.

If any of that overlaps with what you're working on — or if you think a piece of the model is wrong — I'd genuinely like to hear about it.

Get in touch

The fastest way to reach me is email. I read every message, even when replies take a few days.

Email: hello@jobforesight.com

LinkedIn: robiul-islam-consulting

Press, partnership and methodology questions all go to the same inbox. If you're a journalist on a deadline, please say so in the subject line and I'll prioritise.

See your own exposure score

Two minutes, free, no email required to see your headline score.

Take the free 2-min assessment → More about how scores are produced on the methodology page.