Started as a media and journalism graduate who somehow wandered into epidemiology, stayed for the math,
and now builds software and writes national policy.
I started out in newsrooms and somehow ended up modeling cholera outbreaks, then drafting the policies I now have to live under. The throughline is simpler than the CV makes it look: take a messy real-world problem, find the signal in it, and build something that holds up under scrutiny, whether that something is a statistical model, a policy, or a piece of production software.
I work across research, policy, and engineering.
- ποΈ Data & Research Consultant at a2i (ICT Division), focused on digital public infrastructure, interoperability, and digital governance
- π οΈ Founder of Verne, a forward deployed engineering company
- βοΈ Forward Deployed Engineer at Ophanix, where I built the Ophanix Control Panel
- π Quantitative researcher across political science, public health, and food systems
- βοΈ I draft national policy, publish research, and write on system design
Policy. I've helped author some of Bangladesh's core digital policy: lead drafter of the National Startup Policy, committee member on the National AI Policy, drafter of the National Blockchain and Cloud Policy, and co-author of the National Digital Transformation Strategy.
Research. Years of quantitative work across fields that rarely share a CV: predicting cholera outbreaks, detecting heat stress in urban slums, modeling food systems with general-equilibrium methods, and tracking conflict across 23 districts. Different domains, same craft, which is to respect the uncertainty and never report a number you can't defend.
Funding. I've written grants and helped raise research money from the institutions that pay for this kind of work, including the Wellcome Trust, NIH, the Gates Foundation, and Bloomberg.
Words. Peer-reviewed publications (collected on ORCID), opinion columns in the national press, and long technical write-ups on system design.
National policies and strategies I have helped author or draft:
- Lead Drafter, National Startup Policy of Bangladesh
- Committee Member, National AI Policy of Bangladesh
- Drafter, National Blockchain and Cloud Policy
- Co-Author, National Digital Transformation Strategy
- Co-Author, Posts and Telecommunications Division Transformation Strategy 2025β30
Quantitative and computational work spanning several fields:
- Public health: predictive models for cholera early warning, heat stress detection, effective coverage analysis
- Food systems: spatial microsimulation, computable general equilibrium modeling, nutrition-sensitive policy
- Political science: conflict analytics, peace and violence datasets, genocide studies
- Computational social science: NLP on million-document corpora, semantic network analysis of patents
π Publications on ORCID
- π³οΈ Latent-Markov turnout models β treating "I'll probably vote" as a hidden state that drifts between survey day and election day, so a poll becomes an honest range instead of a falsely precise number.
- π Rich-text versioning β Git-style branching, diffing, and merging for documents, built on Merkle trees and CRDTs instead of line-by-line text diffs.
- π§ Geometric deep learning β neural networks for data whose shape carries meaning: graphs, meshes, and other structured objects.
Tools I reach for regularly, grouped by domain. Click an icon to open its docs.
Machine Learning, Deep Learning & NLP
PyTorch |
TensorFlow |
scikit-learn |
Pandas |
NumPy |
SciPy |
Gensim |
Hugging Face |
LangChain |
Streamlit |
PyTorch3D |
LangGraph |
Databases & Research Tools
MySQL |
PostgreSQL |
MongoDB |
Redis |
Git |
Elasticsearch |
Neo4j |
MAXQDA |
KoboToolbox |
NVivo |
I publish notes on system design and write summaries of the papers I'm reading.
- π Research Paper Blog