// data scientist & AI engineer
Data Scientist & AI Engineer
I build things that explain themselves, from LLM-powered career agents to real-time ship trajectory predictors to Sport Analytics dashboards.
3+ years across enterprise analytics, forecasting, and data products. Currently pursuing an MSc in Data & Knowledge Engineering at OvGU, Germany.
// about
Growing up in the Himalayas of Sikkim taught me to look for patterns early. That curiosity eventually took me from computer engineering into data science and now AI engineering.
Today I build data-driven products across analytics, machine learning, and LLM applications. Alongside 3+ years of industry experience, I work as a Research Assistant at OvGU, supporting graduate tutorials in Machine Learning and Data Mining.
// selected work
Projects that show how I turn models, data, and interfaces into usable products.
Featured Project
An interactive professional assistant built with RAG, Groq, and LLaMA that answers questions about my background, skills, and project work in context.
Outcome: Turns a static resume into a recruiter-friendly product demo.
Forecasting & Simulation
Forecasted ship trajectories with LSTM and Transformer-based models using AIS and weather data, exposed as an interactive simulation interface.
Outcome: Combines modeling, weather context, and decision support in one workflow.
Analytics Product
A Streamlit application mapping the geographic talent pipeline of Indian football to surface structural gaps, regional patterns, and scouting opportunities.
Outcome: Shows data storytelling, product thinking, and applied analytics in one project.
Applied NLP
A restaurant review authenticity platform built with HCAI and explainable AI principles to assess sentiment, score reviews, and make moderation more transparent.
Outcome: Demonstrates applied NLP work with user trust and explainability in mind.
OvGU Magdeburg
MSc Thesis — 2025/26
MSc Thesis · In Progress
Researching LLM-as-judge calibration and human agreement for evaluating RAG-based Q&A systems over university course content. Zero-budget local stack with Qwen, Llama 3.1, and Prometheus 2.
Outcome: Explores how to reliably score LLM answers without expensive human annotation at scale.
// career path
I work at the intersection of analytics, machine learning, and product delivery. The through-line: take messy data or complex models and turn them into usable decisions.
Otto Von Guericke University
Support graduate teaching in ML and Data Mining. Contributing to applied research workflows in RAG evaluation.
Data Radian Technologies (Fortune Brands Innovations)
Delivered 50+ analytics dashboards and data workflows for enterprise decision-making across supply chain and marketing teams.
Flex by Finaxar (Part-Time)
Led BI reporting and stakeholder-facing insights for a fintech product, supporting 3 cross-functional teams.
Smart Energy Water
Built predictive models and data pipelines for utility consumption forecasting, improving prediction accuracy by 18%.
// education
2024 – Present
Otto Von Guericke University
2017 – 2021
BML Munjal University
// get in touch
If you're hiring, collaborating, or reviewing my work, send a message and I'll get back to you.
Location
Magdeburg, Germany (Open to Relocation)
Message sent!
Thanks, your message was sent. I usually reply within 1–2 days.
Something went wrong.
Please try again or email me directly.