How it actually went.

I got the idea for LifePath during one of those late-night “what if I'd done things differently” spirals, the kind everyone has, except I couldn't find anywhere that actually answered the question with real data instead of vague Reddit threads or AI guesses dressed up as advice.

So I started building the thing I wished existed. Not a quiz, not a horoscope, an actual framework for comparing life paths using sourced data: what does the average career switch into consulting actually do to your five-year trajectory, what does relocating actually cost versus what people assume it costs. I mapped out how the comparisons should work, built an interactive prototype to pressure-test the logic, and I'm now deep into scaffolding the real version, Next.js on the front end, Supabase underneath.

The honest bit is that most of the hard work so far hasn't been the code, it's been the sourcing. Deciding what counts as “real data” for a question as messy as “should I move to Berlin” is a much harder problem than it sounds, and figuring that out has taught me more about structured decision-making than any lecture has.

MoneyFlow started as my final year group project, but the idea came from something pretty ordinary: staring at my own bank statement, technical background and all, and still not being able to answer “where did my money actually go this month?”

I pitched it as a personal finance tracker built specifically for students, not repurposed banking-app features. Working with two coursemates, I took on the Product Manager / Business Lead role, market research, UI/UX mockups in Figma, writing the user stories, and later authoring the full technical report.

Leo led the engineering: the CSV parsing logic, the database schema, and integrating Mistral 7B (an open-source LLM) to generate the AI insights, running entirely locally so no one's bank data ever left their device. Prabhashini owned testing, validating our CSV parser against real statements from three different banks and stress-testing the AI outputs for edge cases.

The bit I'm proudest of isn't a feature, it's how the team actually worked. I didn't want to just hand over a spec and wait, so from around week 7 I started properly learning the stack myself: Python, Streamlit's reactive model, Pandas, enough SQLAlchemy to understand why Leo's database decisions in week 4 were shaping what we could and couldn't build by week fourteen. That meant I could sit in implementation conversations and actually contribute, not just relay requirements back and forth.

It paid off in the numbers: 87% transaction categorisation accuracy, 14 of 15 usability testers completed the full workflow without help, and AI insights scored 4.1/5 for relevance. We also ran a full refactor near the end, cutting a 1,000+ line monolithic file down to a clean, modular structure, the kind of unglamorous work that doesn't show up in a demo but makes the codebase actually maintainable.

Personally, university has been so much more than just lectures and assignments. It's been about finding my people, building my community, and soaking up every opportunity that came my way ⚡

I somehow went from being the new kid on campus to becoming the Social Secretary for the Pakistani Society 🇵🇰 Planning events, meeting incredible people, and creating spaces where students felt connected quickly became one of the best parts of my uni life 🎒 At the same time, I got back in the ring, sparring and pushing myself for amateur fights when it comes to boxing 🥊 And because I can never sit still, I got involved with the Kent Corporate Society too, connecting with industry professionals early on.

Year 2 is when things started to shift. The society events and networking gave me a taste for building things myself rather than just organising around other people's ideas, so I started teaching myself to code properly and getting into AI, spending less time running events and more time actually building. It wasn't a clean handover; I was still boxing, still showing up to Corporate Society events, but the energy was clearly moving toward “what can I make” rather than “what can I organise.”

Year 3 is where that shift became the main thing. Most of my time outside coursework now goes into building, platforms, prototypes, actual products rather than ideas on paper. It's a different kind of challenge to running a society: less about people showing up on the night, more about whether the thing you built actually works and solves something real. Looking back at the three years, it's a strange arc, from Social Secretary to boxer to builder, but each one taught me something the next one needed.

Over the summer, I joined the ODI (Outcome Delivery Incentive) Excellence team at Severn Trent, where I dove into a bunch of projects all centred around performance, insight, and continuous improvement.

My days were a mix of deep-diving into flooding data, mapping out reasons behind customer complaints, building dashboards, and supporting internal audits.

I worked on visual trackers, designed newsletter content, and helped make sense of complex data in a way that people could actually act on.

One of my biggest takeaways? Turning numbers into stories that matter to different teams.

I also picked up a lot on the go, how to pull together data from different systems, what makes a dashboard useful (not just pretty), and how to present insights that actually get listened to.

Outside the typical analyst tasks, I made a conscious effort to explore the company from every angle.

I organised a full site visit to Strongford (a net zero site!) for interns to see operations up close, and teamed up with HR to host a leadership panel with senior speakers, including Liv, where we created space for open, honest conversations about career journeys and resilience.

I also jumped at the chance to shadow different teams, like spending a day with the recruitment team to understand how Severn Trent handles thousands of applications (spoiler: most of it is manual!).

I reached out to the Power BI team, booked time with an internal mentor, and started building my first dashboard.

And to give back, I volunteered to run mock interviews for Year 12 students and took part in a local recruitment drive.

This year, I've made sure I step out of my comfort zone and build connections with professionals across different industries.

I've attended incredible networking events, from PwC's “Meet the Middle East Team” to JP Morgan's “Insights Into Asia” networking session. I had the chance to listen to Simon Squibb share his journey as an entrepreneur and connect with like-minded peers afterwards. Attending an event hosted by Neuberger Berman exposed me to the world of asset management, something that has really sparked my interest for a potential career path 👀

Each experience has taught me the importance of staying curious and building authentic relationships and skills that I'll carry forward as I grow in my career. ✨