Open to opportunities

Yamuna Mandala

@yamunamandala

Message

I turn messy data into trustworthy analysis and validated ML pipeline outputs.

India

Message

What I'm looking for

I'm looking for an analysis-focused role where I can turn messy data into trusted outputs, validate pipeline quality, and contribute to products teams ship. I value remote collaboration, self-directed learning, and opportunities to grow with modern data tools.

I've cleaned, preprocessed, and encoded the UCI Adult Census dataset during IBM's AI & ML Virtual Internship, building a Random Forest model that reached 87% accuracy. I also documented model strengths and failure points while handling malformed and unexpected data gracefully.

At BCG X's GenAI Job Simulation, I extracted and validated structured financial data from SEC 10-K and 10-Q filings with an NLP extraction pipeline. I tested REST API workflows end to end to verify reliable outputs before use.

In my major project, I built a Python, Whisper, NLP, and Flask/Streamlit pipeline that turns raw meeting audio into structured minutes, benchmarking language-classification accuracy at 86.67% across a held-out 60-recording test set.

Experience

Work history, roles, and key accomplishments

AI & ML Virtual Internship

IBM

Jun 2025 - Jul 2025 (1 month)

Worked fully async and remote through the complete data pipeline, including cleaning, preprocessing, and feature encoding, on the UCI Adult Census dataset, producing a model that reached 87% accuracy with Random Forest. Documented pipeline successes and failures, applying structured exception handling to manage malformed or unexpected data.