Data Scientist & ML Engineer

Muntazir
Ali Mughal

MSc Data Science · King's College London

Building end-to-end ML systems — from raw data pipelines to deployed models with explainable outputs. Available from mid-June 2026.

Scroll
About

Building things that
actually work

I'm an MSc Data Science student at King's College London with a First Class BSc in Software Engineering from Liverpool John Moores University. I care about models that perform, pipelines that run in production, and outputs that non-technical stakeholders can actually use.

I've built financial risk platforms grounded in SEC 10-K filings, clinical readmission models with SHAP explainability deployed on Streamlit Cloud, real-time cognitive load estimators from keystroke patterns, and ML pipelines from physical sensor hardware to production inference.

I target roles where engineering rigour meets data science — not just notebooks, but systems that ship.

101K+
Patient records modelled
0.63
AUC — clinical benchmark
500+
SEC filings processed
Financial Risk · RAG · NLP
FinRisk Terminal
AI-powered financial risk intelligence platform. XGBoost risk scoring on 10 S&P 500 companies backed by 500+ SEC 10-K filings. Groq-powered RAG copilot deployed on Streamlit.
PythonDuckDBXGBoostLangChainGroqChromaDB
View on GitHub ↗
Healthcare · ML · Interpretability
Hospital Readmission
End-to-end readmission risk predictor on 101,766 diabetic patient records. XGBoost + SMOTE + SHAP waterfall explanations. Deployed live on Streamlit Cloud.
XGBoostSHAPSMOTEStreamlitscikit-learn
Live Demo ↗
Behavioural AI · Chrome Extension
CogniType
Real-time cognitive load estimator from typing patterns alone — no wearables, no cameras. Chrome MV3 extension feeds a FastAPI backend with keystroke batch processing.
FastAPISQLiteChrome MV3PyTorchPython
View on GitHub ↗
Hardware ML · Signal Processing
Sign Language Glove
Flex sensor glove translating 28 ASL gesture classes to text in real time. 5K+ labelled sensor samples, 95%+ accuracy with scikit-learn classifiers.
scikit-learnSignal ProcessingEmbedded CPython
View on GitHub ↗
Selected Work

Projects

Financial Risk · RAG
FinRisk Terminal
XGBoost risk scoring + Groq RAG copilot grounded in 500+ SEC 10-K filings.
PythonDuckDBLangChainGroq
GitHub ↗
Healthcare · ML
Hospital Readmission
101K patient records, SHAP explanations, live on Streamlit Cloud.
XGBoostSHAPStreamlit
Live Demo ↗
Behavioural AI
CogniType
Cognitive load from typing patterns. Chrome extension + FastAPI backend.
FastAPIChrome MV3PyTorch
GitHub ↗
Hardware ML
Sign Language Glove
28 ASL gesture classes, 95%+ accuracy from flex sensor data.
scikit-learnEmbedded C
GitHub ↗
Selected Work

Projects

Python
SQL
PyTorch
XGBoost
LangChain
RAG
FastAPI
Streamlit
scikit-learn
SHAP
ChromaDB
Git
MLflow
DuckDB
Docker
Capabilities

Skills

Languages
PythonSQLJavaScript
ML / AI
PyTorchscikit-learnXGBoostSHAPHuggingFace
LLM / RAG
LangChainChromaDBGroqOpenAI
Engineering
FastAPIDockerGitMLflowDuckDB
Capabilities

Skills

Background

Education &
Experience

2025 —
Present
MSc Data Science
King's College London
Statistical learning · ML systems · Data engineering · Applied AI
2022 —
2025
BSc (Hons) Software Engineering — First Class
Liverpool John Moores University
Full-stack systems · Databases · Applied ML · Sign Language Glove dissertation
2021 —
2022
BTEC Foundation Diploma
Pearson
Embedded systems · Microcontrollers · Electronics
Get in touch

Let's build something
worth building

Open to data scientist, ML engineer, and data analyst roles in UK Tech & SaaS.
London, United Kingdom.