Data Scientist · ML Researcher · LLM/NLP · Liverpool, UK

Shweta Debjit Sarkar

I'm a data scientist and machine learning researcher who spends most days elbow deep in messy real world data clinical time series, environmental exposure records, behavioural cohorts and increasingly, in language itself: building NLP pipelines and exploring what LLMs actually understand versus what they're good at sounding like they understand. I care about models that are accurate, explainable, and a little kinder to the world they touch. Currently applying to PhD programmes.

About me

the (tidy) corner where I keep my models honest

I am a data scientist and machine learning researcher with practical experience building models on complex, multi source datasets including environmental exposure data, clinical time-series, and large scale behavioural cohorts.

Across four independent research preprints, I have developed strong instincts for making models not just accurate but interpretable and actionable. My work spans spatial epidemiology, clinical risk prediction, and ICU patient monitoring and increasingly, NLP and large language models, from retrieval augmented pipelines over clinical literature to auditing bias in real hiring language.

I am drawn to doctoral research that applies these skills to sustainability and systems level problems building tools that help researchers and policymakers see complex interactions in real time, not just in reports.

Recent publications

output, newest first — tap a card to open the record

2026

Cervical Cancer Risk Prediction

SHAP-driven clinical risk stratification with class imbalance handling.

Tap to expand →
2026

Student Depression Risk Index

A mathematical index for personalised behavioural guidance.

Tap to expand →
2026

ICU Hypotension Prediction

Explainable early prediction using routine vital signs.

Tap to expand →
2026

Stochastic Imputation Method

Local bounded imputation for ordered numerical sequences.

Tap to expand →

Applied research

some more successful than others — tap a card for the full story

2026

Satellite Wildfire Classifier

99.5% accuracy — but was it looking at the right thing?

Tap to expand →
2026

Coral Reef Health Classifier

VGG16 vs EfficientNetB0 on 923 images.

Tap to expand →
2026

Skin Lesion Classifier

Scratch CNNs vs ResNet50 transfer learning.

Tap to expand →
2026

Job Posting Bias Analyser

123,842 postings, one lexicon, some uncomfortable truths.

Tap to expand →
2025

Wildlife Camera Trap Detector

Teaching a model to actually look at the animal.

Tap to expand →
2026

AI Agency in Student Learning

A classroom study on copying vs understanding.

Tap to expand →
2026

Clinical RAG

Parkinson's & Alzheimer's literature, made queryable.

Tap to expand →
2026

T2DM Metabolomics Analysis

13 urinary biomarkers, cross-validated AUC 0.985.

Tap to expand →
2026

Air Pollution & Breast Cancer

Spatial epidemiology across English regions.

Tap to expand →
2025

Rainfall Classification

Atmospheric data, binary prediction.

Tap to expand →
2026

Breast Cancer Wisconsin

XGBoost, SHAP, and a clean AUC-ROC.

Tap to expand →
2025

MoovBuddy Cohort Analysis

100,000+ users, 88% drop-off, one MSc thesis.

Tap to expand →

Experience & education

the paper trail

Teaching Assistant — Maths, Science & Computing
Jan 2026 – Present
Dixons Fazakerley Academy · Liverpool, UK
  • Supported classroom delivery across Maths, Science, Computing, and Business for GCSE and A-Level students.
  • Delivered targeted interventions for students with autism, ADHD, and dyslexia.
  • Collaborated with senior teachers on personalised study plans.
Research Intern — User Behaviour Analysis (MSc Thesis)
May – Aug 2025
MoovBuddy Ltd · London, UK
  • Longitudinal cohort analysis on 100,000+ users across 10+ countries.
  • Identified 88% install-to-subscription drop-off; modelled country-level variation.
  • Built a four-phase recommendation framework for the marketing team.
Teaching Assistant — Maths, Science & Computing
Sep – Dec 2025
Childwall Abbey School · Liverpool, UK
  • Supported students with mild to severe neurological disorders.
  • Assisted coursework delivery and exam preparation.
Business Analyst
Dec 2021 – Mar 2023
Indothai Group · India
  • Analysed 300+ daily transactions using SQL and Excel.
  • Built dashboards for service quality and revenue trends.
MSc Business Analytics & International Business
University of Dundee
Sep 2024 – Sep 2025 · Dundee, UK
Machine Learning · Statistical Methods · Big Data Analytics & Predictive Modelling · Database Management · Deep Learning
🏆 Postgraduate Scholarship · 🥈 2nd Place, First Semester Group Projects
BEng Electrical Engineering
S.B. Jain Institute of Technology, Management & Research
Aug 2015 – Aug 2019 · Nagpur, India
Signal Processing · Applied Mathematics · Probability and Optimisation · Network Analysis · Control Systems
🎖️ General Secretary, Inter-Departmental Forum

The toolkit

what's actually in the toolbox

Programming & Databases

Python (Advanced), R, SQL

ML & Statistical Methods

XGBoost, Random Forest, SVM, LSTM, Autoencoders, SHAP, SMOTE/ADASYN, Survival Analysis, MICE/KNN Imputation

Spatial & Data Analysis

GeoPandas, Spatial Data Linkage, Epidemiological Methods, Exposure Analysis

ML Frameworks & NLP/LLM

TensorFlow, PyTorch, Scikit-learn, Hugging Face Transformers, Retrieval-Augmented Generation (RAG), NumPy, Pandas, SciPy

Visualisation & Dashboards

Matplotlib, Seaborn, Power BI, Amplitude

Tools & Platforms

Jupyter, Git/GitHub, Google Colab, Docker, Excel (Advanced)

Let's collaborate

say hello — I don't bite (probably)