I'm Mohammed Imran. Founding PM twice, program lead at Uber, and the person who'd rather build the prototype than book another meeting about it.
I make messy things run. Then I make them run without me.
a year in projected savings, from an automation roadmap for Uber's Global Field Operations
conversion rate as founding PM at Aastey, lifting monthly revenue 55%
less manual QA review, with an AI quality platform built at Uber
of the team actively using AI, through a three-tier adoption program
Engineer, founding PM, MBA, program lead, AI builder, Batman. Okay, maybe not the last one. The hats change, the habit doesn't: find the messy problem, get the right people around it, and ship what works. Then make it better every day. Call it ambition. I call it being a goal digger.
My instinct is good. I still check it against the numbers before I decide.
When things break, someone has to stay steady. I'm usually that person.
I love to win. Helping the team win feels better.
Every program and product decision starts with who it's for.
Small improvements, every single day.
Ambitious about outcomes, and not shy about it.
Programs and products I led, and what each one changed. Click any project for the full story.
Two startups, built from zero. One made it to Shark Tank India.
I build in public, so I can lead AI products from the inside, not just write the roadmap for them.
An AI agent that reads a pull request, understands the change, and leaves review comments like a careful senior engineer.
Follow the build on LinkedInPlugs GitHub into any AI assistant as tools it can call: read issues, open pull requests, search code.
Watch it on GitHubGrowth, valuation and market-entry strategy for global brands. One of them won an award.
Product innovation and market positioning for Kenwood's home appliance range. Won the Chartered Institute of Marketing award for Marketing & Strategy.
Award-winning teamValuation, financial analysis and investment risk on Mercedes-Benz (MBG), ending in a clear call.
Equity researchWhy the UAE drew $30.7B in foreign direct investment in 2023, ranked 2nd in the world, and where investors should play next.
Macro strategyEvaluated 60+ startups with financial models and market research to build a deal pipeline, plus an AI investment thesis for the UAE and India.
PwC, Venture
Go-to-market and Shopify growth strategy for Kayali, plus GTM work for GravityiLabs and a strategy dashboard MVP for Apparel Group.
PwC, GTM
Strategic audit and competitive benchmark, with a plan for $2M in savings and 40% revenue growth.
Consulting
General Mills UAE market entry
Jumeirah brand audit
Tesla innovation strategy
Nestlé brand expansion
Engineer first, then founding PM, then business school, then programs at global scale.
Vendor programs on Uber's Merchant Global team.
Led 5 transformation programs across Global Field Operations.
Store revamp: 1.8x conversion. Regional launches across the Middle East and Africa.
Evaluated 60+ startups for the deal pipeline.
8.5x conversion. Featured on Shark Tank India.
MRR up 45%, retention to 72%.
Pitched an automation that saved Discover Bank's ops team about 256 hours.
GPA 3.92 / 4.
The MERN stack behind my side projects, and why I still read the code.
CGPA 8.2 / 10. Co-founded SureLocal, a 24/7 medicine delivery startup: about 200 users, and the first startup in SRM's incubator to raise seed funding.
Strategic Communication online
Published on Medium
I'll bring the framework. You bring the chaos.
Building something, hiring, or just want to swap notes on AI and ops? I usually reply within a day.
An AI layer that reads every support conversation and routes each issue to the team that can fix it.
QA teams could only review a small sample by hand, and CSAT scores hid real issues. Problems surfaced late, and to the wrong teams.
Built an AI layer that flags issues in four lanes and routes each one to the team that owns the fix. Recognised by leadership for ownership.
Uber's data rules ruled out OpenAI, so the platform runs on Google tools and Uber's in-house Michelangelo platform. That one constraint shaped the whole design. Internally, it was nicknamed Cringebot.
Click a lane to see what it does.
Spots product bugs and gaps hiding in conversations and flags them to the owning team, so fixes happen upstream instead of one ticket at a time.
Plenty of manual work, no shared way to decide what to automate first. So I built one.
Every market had manual workflows it wanted automated, and every team had its own favourite. Without a shared yardstick, the loudest request won instead of the most valuable one.
Every workflow lands in one box.
Access to AI tools is easy. Getting people to change how they work is the hard part.
Tap a tool to see where it fits.
Everyday writing, summarising and analysis, inside the tools people already use.
Two quiet fixes that made Uber's field operations fairer, cheaper and easier to measure.
Refunds and goodwill credits depended on each agent's judgement, so two similar customers could get very different outcomes. I replaced that with risk-based compensation tiers, so the response fits the situation and costs stay predictable.
Every market named the same teams differently, which made comparing, staffing and reporting painful. A single taxonomy across Global Field Operations cut reporting errors and gave leadership consistent analytics.
A full redesign of Sony's Middle East online store, from the first click to checkout.
Launched Headphones, Reon and Alpha camera campaigns across the Middle East and Africa, adapting global campaigns with Post Malone, Jennifer Lopez and World Aquatics. Wrote the headphones launch strategy, built around a two-wave influencer plan. The campaigns drove 500K+ site visits.
India folded 29 labour laws into four codes covering roughly 500 million workers. I built a system that answers plain-English questions about them with cited passages, then measured how often it actually finds the right text.
Most RAG demos stop at "it answered my question." I wrote a 32-question answer key, scored every change against it, and kept a log of what worked and what didn't. Every number here reproduces from a clean clone on a fresh machine.
I also built a quality gate. It caught that the Code on Wages download was a cover document, not the code itself, and kept it out instead of letting bad data in silently.
Click each step.
Download the codes, strip boilerplate, split into 1,000-character passages with 150 characters of overlap.
Pick a measure.
The failures taught me as much as the wins.
Being honest about the gap is part of the job.
"Can my boss fire me without any warning at all?"
Passages from the safety code about fire and flooding in mines. It matched the word, not the meaning.
Industrial Relations Code, section 70: one month's written notice before retrenchment.