An accountant who learned to build the systems he wished existed.
I started my career in accounting — reconciling ledgers, closing books, and building spreadsheets that nobody else wanted to touch. Somewhere between the third pivot table and the seventh manual month-end, I decided the tools had to be better. That's when I picked up Python. Today, I build ERPs, dashboards, and research pipelines that quietly replace hundreds of hours of manual work.
- Base: Bangladesh · open to remote & relocation
- Domains: Finance · ERP · BI · Tourism research
- Core stack: Python, FastAPI, PostgreSQL, Streamlit
- Research focus: Digital visitor economy & GLMs
- Availability: Remote roles · freelance · PhD 2026
My journey
My background is in accounting and finance — the language of debits, credits, and quiet spreadsheets that keep organizations running. What I quickly noticed, however, was that most finance teams were doing sophisticated work with primitive tools. Reports were manual. Reconciliations were fragile. Insight was buried under formatting.
So I taught myself Python. Then SQL. Then FastAPI, NiceGUI, PySide6, and everything in between. What started as automating my own tasks turned into designing ERP systems from scratch — double-entry accounting engines, payroll modules, procurement workflows, and executive dashboards — for real businesses.
Where I'm going
The next chapter is academic. I'm actively preparing to pursue a PhD in Tourism Analytics, with a research focus on the digital visitor economy, destination branding, risk and crisis modelling, and sustainability metrics. My working papers combine time-series forecasting, generalized linear models, and large-scale survey analysis — using R and Python as the primary tools.
What I care about
I care about work that is measurable, reproducible, and honest. I care about dashboards that people actually open. I care about research that doesn't hide behind jargon. And I care about building software that a finance director, a data scientist, and a PhD supervisor can each look at and immediately trust.
Principles I refuse to compromise on.
Precision
Numbers matter. A rounding error in row 40,000 is still a lie. I ship code and reports with the same audit-grade discipline I learned in accounting.
Reproducibility
Every analysis is a pipeline — versioned, documented, and re-runnable. If the result can't be reproduced, it isn't a result.
Clarity
Executives don't want models — they want decisions. Every chart, table, and API response should answer a question a human actually asked.