Analyst · Researcher · Consultant
MSc Business Analytics (Distinction) · University of Surrey
Accepted — IFORS 2026 International Conference · Graduate Visa, full UK work rights
I build quantitative models for policy and operational problems, working across Bayesian inference, discrete-event simulation, efficiency analysis, and NLP. Coming from a Finance background, my research spans misinformation dynamics, fuel poverty risk, and infrastructure efficiency in the Global South. I am looking for data and analytics roles in the UK, and for research pathways in operational research and development economics.
Accepted for oral presentation at the 24th IFORS Conference, Vienna, July 2026. Frames social media content moderation as a resource-allocation problem under partial observability. Combines NLP, transformer-based sentiment modelling, Markov-chain transition dynamics, and Monte Carlo robustness testing to construct two composite indices for narrative-level intervention prioritisation.
Assembles 2.6 million UK property records from six government sources. Applies MCMC Bayesian inference to generate household-level fuel poverty probability distributions from aggregate LILEE statistics. Validates using Random Forest (0.88 ROC AUC, Pipeline B) and XGBoost across dual pipelines isolating signal leakage. Submitted to Energy Policy (Elsevier).
Applies Data Envelopment Analysis to assess public infrastructure efficiency across Global South economies, examining whether constraints on infrastructure development are primarily financial or institutional. Work in progress.
M/M/c queueing theory and discrete-event simulation in Simul8 to model a UK forensics DNA pipeline. Five operational scenarios evaluated across 500 trials. ROI analysis using Monte Carlo simulation via Palisade @Risk.
Efficiency AnalysisBenchmarks 130 publicly owned US electricity distributors using Data Envelopment Analysis (PIM-DEA) and Stochastic Frontier Analysis. Translog cost function estimation with COLS, MOLS, and SFA under VRS and CRS assumptions.
Predictive ModellingPredicts accident severity from 2,400 records using SAS Viya. Addresses severe class imbalance via bootstrapped balancing. Bayesian Network achieves 3.04 cumulative lift at top decile with 70.7% sensitivity.
Bangladesh's banking sector carries over BDT 1.56 trillion in classified loans. This piece traces the structural roots of that crisis — from 1972 nationalisation through to regulatory capture and directed lending.
MacroeconomicsHow post-pandemic US monetary tightening, the Ukraine shock, and Bangladesh's own reserve management failures combined to produce a foreign exchange crisis.
MacroeconomicsExamines the structural drivers behind de-dollarisation momentum — Fed rate hikes, BRICS currency ambitions, and where Bangladesh sits in the shifting global reserve landscape.
Statistics & Econometrics · Machine Learning/AI & Visualisations · Data Mining & Text Analytics · Performance Analytics · Equity Investment Analysis · Operational Analytics · Supply Chain Analytics
Financial Derivatives & Risk Management · Security Analysis & Portfolio Management · Investment Banking · Central Banking · Credit Management · Econometrics · Monetary & Fiscal Policy