The newest person
in the room.

I grew up in Dinajpur, in the north of Bangladesh, in a household that treated reading as the default answer to boredom. If you were restless, you found a book. If you had finished the book, you found another one.

It was a small world, and I mean that kindly. Everyone knew everyone’s rhythms. Days had a shape. Nothing about it was preparing me for military school, for China, or for a career in finance, which is probably why it matters so much to me now. It’s the one chapter where I wasn’t new to anything.

What Dinajpur gave me was a habit: watch closely, keep the details, and don’t confuse noise for substance. Almost every room I’ve walked into since has been louder than that house. The habit still works.

At twelve I left home for Rangpur Cadet College, a military school. On a map it is sixty-six kilometres from Dinajpur. At twelve, it might as well have been another country.

It was the first room I had to learn by myself. Routine, repetition, hierarchy, accountability: the rules were written down. How the place actually worked was not, and nobody explains that part. You watch the people who have been there longer, you notice what they notice, and you adjust.

The structure was external at first. Somewhere along the way it moved inside. I learned to keep my own rhythm when the environment was demanding, to stay composed when things felt uncertain, and to keep showing up after the work stopped being exciting. Effort, it turns out, has a texture. Most of it is quiet.

I don’t romanticize every hard place. But this one built the floor that everything else stands on.

At sixteen I left again, this time for a city in China I had never seen, on a full scholarship to United World College Changshu. I didn’t speak the language. Neither did a lot of my classmates, who came from more than a hundred countries.

That was the point of the place. UWC puts people from very different realities in the same classrooms and the same late-night conversations, and asks them to find common language. You learn quickly that smart people can start from completely different assumptions, and that the fastest way across is curiosity plus honesty.

The International Baccalaureate rewarded range. You carried several disciplines at once and were expected to connect them rather than keep them apart. I still work that way. Whether the problem is a product, a regulatory filing or a client’s workflow, I look for the bridge before the silo.

If Rangpur taught me how to survive a new room, Changshu taught me something better. You don’t just endure the unfamiliar. Done well, it makes you larger.

University took me to Luther College in Decorah, Iowa, a town of about eight thousand people. Winter there teaches patience. A small campus teaches you that depth beats breadth, if you let it.

I majored in Data Science with a minor in Computer Science and graduated in 2022. The degree gave me tools, but the bigger change was in the questions I asked. Where did this information come from? What is signal and what is noise? What decision is it actually supposed to support?

Junior year, I applied to Blackstone. Or I thought I did. When the interview invitation arrived, I read the name three times. BlackRock. I had prepared private equity case studies and was about to walk into a conversation about fixed income risk. I learned what duration meant very quickly.

That mix-up sent me down a rabbit hole I have never fully climbed out of: how money is actually organized, and why two firms with nearly identical names ended up as opposites. I eventually wrote it all down in a Field Note.

My first serious data work was at Mayo Clinic in Rochester, Minnesota, as an AI intern. The datasets were clinical, the users were research teams, and the distance between an analysis and a consequence was short.

Two lessons stuck. The first was that reliability beats cleverness. A reporting pipeline that works every single morning is worth more than a model that impresses once. The second was that messy, real-world data isn’t an obstacle to the job. It is the job.

The part I didn’t expect to enjoy was the translation. Clinicians brought questions that were important and vague at the same time. Turning them into something measurable meant learning how they thought first. It was the first time the skill I had been practising since twelve showed up in a job description.

SkyBridge was my first close look at how institutional money decides where to go. As a data scientist supporting manager selection, I helped build quantitative screens across more than 200 hedge fund profiles, and dashboards that tracked exposure by strategy, geography and liquidity.

The models mattered less than I expected. What mattered was whether someone could read the output in a meeting and decide something. I contributed to investment memos, and learned that a good memo isn’t a pile of findings. It’s an argument, shaped to fit how decisions actually get made.

That’s where I stopped thinking of narrative as decoration. Narrative is translation: complexity, turned into something a decision-maker can act on with confidence.

After graduating, I joined BlackRock as an analyst in Data and AI, working in Wilmington, Delaware, with the Aladdin platform close at hand.

Yes, the same BlackRock I once confused with Blackstone.

I worked on risk reporting and fund performance metrics, built dashboards connecting capital-flow analytics to strategy performance, and presented what I found to senior stakeholders. The lesson I carried out of the building: a trillion dollars is a spreadsheet problem, a risk problem and a storytelling problem, all at once. You don’t get to solve only one of them.

BlackRock also changed my taste. Up close, the best financial infrastructure isn’t impressive. It’s dependable. It holds up, it scales, and it makes complex work legible without pretending the complexity isn’t there. I came away caring less about novelty and more about trust, which is most of what finance software is really selling.

In 2023 I joined Dasseti, a fintech company that builds due diligence software for investors. My title is solutions engineer. I’m not a software engineer, and that distinction is the whole job.

A solutions engineer sits between two groups who both know exactly what they mean and rarely mean the same thing. On one side are investment and operations teams, who need to evaluate managers, collect data and keep a record they can defend. On the other side is the product, which can do a great deal if someone explains the problem to it properly. I learn how a client’s team actually works, then translate in both directions: what they need into what the product can do, and what the product does into their language.

I wanted a startup because I wanted to be closer to the user, and proximity turns out to be clarifying. When you’re in the room with the person doing the work, vague thinking gets exposed quickly. So does the difference between impressive language and something that actually gives someone their afternoon back.

The twelve-year-old working out the unwritten rules at Rangpur was practising for this. I just didn’t know it had a name.

These days I live in Land O’ Lakes, Florida and work remotely. The day job is Dasseti. The rest of my time goes to three things that turn out to be the same habit, pointed in different directions.

Field Notes. Long-form pieces on how capital actually moves: GameStop’s bid for eBay, Uber’s $32.8 billion hole, the 56,000-plus advisers inside Form ADV. Every big number has a story behind it. I go and find it.

The Garage. Small tools I wanted and couldn’t find, from a cricket fielding planner to a way to read an investment adviser’s entire regulatory file in one place. I build them with AI coding tools, which rewards the same skill as the day job: describing exactly what should exist, then refusing to stop until it does.

Cricket. In Florida, I play league cricket as an all-rounder who bowls right-arm fast, and write up the close finishes. Seven runs to defend in the last over. Four wickets and a chase. I still replay them.

Numbers into stories, problems into tools, matches into memory. It’s all translation.