I built a tool to fix a broken system I hated watching my friends suffer through, only to realize the co-founder I needed to make it real was sitting right across our kitchen table.
During my MSc in Management at Durham, I was sitting in a lecture on optimizing AI for productivity when everything suddenly clicked.
For years, I'd watched my friends go through the absolute meat grinder of the graduate application cycle. I saw the exhausted, late-night panic of tailoring CVs, the crushing radio silence of rejections, and the quiet dread of people accepting offers only to realize the start date clashed with graduation. Worse, I watched people land jobs that looked incredible on paper, only to realize on day one that the role was nothing like the description. It felt insane to me that we were making some of the biggest commitments of our early careers based on a vague job listing, a rushed 30-minute interview, and a prayer.
RoleSift wasn't born out of a grand business plan; it started as a messy side project. I just wanted to use what I was learning to make applying for a job feel less like Russian roulette.
But then I started showing it to the people around me — the friends losing sleep over deadlines, the coursemates staring blankly at offer letters they weren't sure they should sign. Every single person I talked to had the exact same look of exhaustion. The frustration wasn't just common; it was universal. There was a massive, painful gap in the process, and nobody was fixing it.
That's the moment the imposter syndrome cleared, and I realized this couldn't just be a folder on my laptop anymore. It had to be real.
But I also knew I couldn't carry it alone. I needed someone to challenge me, pull me out of my own head, and bring the skills I completely lacked. As it turned out, I didn't have to look far at all. I found all of that in Irfan — who I happened to be sharing a flat with. What started as casual venting over morning coffees in our kitchen quickly turned into late-night mapping sessions at the dining table, realizing we were both looking at the exact same problem from different angles.