Skip to content
View thrisharajkumar's full-sized avatar
🎯
Focusing
🎯
Focusing

Block or report thrisharajkumar

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
thrisharajkumar/README.md

Typing SVG

✨ Thrisha Rajkumar

MSc Data Science @ University of Bristol (Russell Group, 2025) | AI/ML Research Support – NIHR-funded NHS Stroke Rehab | Ex-Unilever Data Science Intern | End-to-end ML & AI pipelines with measurable impact

Data Scientist & ML Engineer building production-grade, cloud-native ML systems: data pipelines, feature engineering, orchestration (Airflow), deployment, monitoring, drift detection, and governance. Hands-on across healthcare, industrial, and commercial domains — leveraging predictive modelling, time-series forecasting, NLP, deep learning, generative AI, and XAI (SHAP/LIME).

Tech Stack: Python, SQL, TensorFlow/Keras, Scikit-learn, Airflow, Docker, Azure/AWS, CI/CD, Flutter integration.
Impact: Process efficiency gains up to 40%, energy savings 12–15%, decision-ready clinical and operational insights.

📍 London-based | UK Graduate Visa eligible | Open to hybrid/remote roles in Data Science, ML/AI Engineering, and MLOps

LinkedIn


👋 About Me

Passionate about AI for social good — developing production-grade AI for NHS stroke rehabilitation with personalized exercise recommendations, LLM pipelines, generative models, and user-centered co-design. Bridging ML research, software engineering, and clinical impact, with prior experience in cross-functional analytics at Unilever.


🎓 Education

  • MSc Data Science, University of Bristol (2024–2025) – Distinction track (AI, NLP, Bayesian methods, visual analytics)
  • BTech Computer Science Engineering, DIT University (2020–2024) – First Class Honours (85.4%)

📈 GitHub Activity

Profile Views Summary Stats


💼 Key Experience

AI/ML Research Support & Software Engineer – NIHR Stroke Rehabilitation Project, University of Bristol (Apr 2025–Present)

  • Led development of NIHR-funded stroke-rehab app, integrating predictive personalization, Clinical-T5 NLP, diffusion/generative models, and SHAP/LIME explainability.
  • Designed GDPR-compliant, cloud-native data pipelines with Python, SQL, Airflow, Azure/AWS; ensured production-ready ML workflows with monitoring and drift detection.
  • Conducted user-centered co-design workshops with patients and clinicians; translated feedback into actionable model improvements.
  • Lead author of HEALTHINF 2025 paper presenting applied clinical AI and production-ready ML pipelines.

Data Science & Analytics Intern – Unilever (Hindustan Unilever Ltd., 2024)

  • Delivered 14+ analytics projects across supply chain, manufacturing, and safety: Python optimisation for 250+ SKUs (40% changeover reduction), ML forecasting (15–20% efficiency gains), energy dashboards (12–15% usage reduction).
  • Automated supply chain and energy-management workflows using Python, SAP, Power BI/Tableau; consolidated 29K+ safety logs (71% reduction).
  • Built scenario- and resource-optimisation models to support staffing, cost, and operational decisions.

Other Experience: Sapienza University research (protein residue prediction), Salesforce/Infosys internships (RPA, NLP), Flutter full-stack projects


📄 Publications

  • Lead Author: “Personalised Stroke Rehabilitation: An AI Pipeline for Exercise Programmes Using a Co-Designed Decision Support Tablet Application” — HEALTHINF 2025 (peer-reviewed, NIHR-backed)

📊 Featured Projects

  • NHS Stroke Rehab AI App – Flutter + ML personalization, clinician/patient co-design, production deployment
  • EMI Surrogate Modelling – Physics-informed neural nets, multi-head attention, R² >0.95, SHAP analysis (thesis)
  • UK Census Visual Analytics – Bayesian imputation, UMAP/t-SNE/KMeans, interactive Tableau dashboards
  • Peri-Operative Time-Series Forecasting – LSTMs/CNNs/Transformers, distributed GPU scaling
  • Protein Residue Prediction (Sapienza) – CIRNet + feature engineering → 78% accuracy

🔹 Core Skills

Python TensorFlow Scikit--learn SQL Airflow AWS Docker Tableau Flutter Generative AI


🏅 Awards & Certifications

  • McKinsey Forward Program (2025)
  • Bristol PLUS Award – Leadership & Employability
  • BILT Student Research Festival – Highly Commended & Funding
  • AWS Academy Graduate – Cloud & ML
  • Infosys AI/NLP/Deep Learning/RPA
  • Salesforce Superbadges

🎤 Talks & Involvement

  • NIHR ABI Dissemination Conference, London (2025)
  • HDRN Workshop (2025)
  • ETH Oxford Hackathon Participant (2025)
  • Postgraduate Representative – Bristol Data Science Society

🎭 Beyond Code

Professional Bharatanatyam dancer | Carnatic vocalist | Marathon runner — discipline, creativity & resilience fuel my work

Open to graduate/junior-mid roles in Data Science • ML/AI Engineering • MLOps — healthcare, fintech, or impact-driven tech
Connect: LinkedIn | thrisharajkumar5@gmail.com

Pinned Loading

  1. FINAL_DATA_SCIENCE_PROJECT FINAL_DATA_SCIENCE_PROJECT Public

    Machine learning based models of wide bandgap generated electromagnetic interference (EMI) in electric vehicle (EV) power electronic converters

    Jupyter Notebook 1

  2. MSc-SocioEconomic-Analytics-UKCensus-BayesianModelling-TableauDashboards MSc-SocioEconomic-Analytics-UKCensus-BayesianModelling-TableauDashboards Public

    Analyzed and visualized changes in education, employment, occupation, and immigration across 348 UK Local Authority Districts using 2011 and 2021 census data.

    Jupyter Notebook 1

  3. Statistical-Computing-Empirical-Methods-Using-R-Programming Statistical-Computing-Empirical-Methods-Using-R-Programming Public

    This repository contains coursework for the Statistical Computing & Empirical Methods unit, which blends foundational statistical techniques, empirical data exploration, and predictive modeling usi…

    1

  4. AWS-WordFreq-Project-Architecting-and-Scaling-a-Cloud-Native-Text-Analyzer AWS-WordFreq-Project-Architecting-and-Scaling-a-Cloud-Native-Text-Analyzer Public

    This repository contains the implementation and documentation for the WordFreq Application, a scalable, cloud-native text analysis pipeline developed for the Large-Scale Data Engineering (LSDE) cou…

    1