Headline

Data Scientist | Generative AI · LLMs · RAG · LSTM · Financial ML | Automated Report Summarization from Hours to Minutes | Open to Full-Time Roles

About

Cutting manual financial analysis time by 40% — that's what an automated ETL pipeline looks like in production. At NUANCE, I build the systems that turn millions of unstructured trading logs into clean, decision-ready data, and the models that turn that data into an edge. My work sits at the intersection of machine learning, generative AI, and financial data engineering. I design and deploy time-series forecasting models — LSTM, ARIMA, SARIMAX — that improve trading trend predictions. I build classification pipelines that surface high-probability trade signals. I've integrated LLMs with Retrieval-Augmented Generation (RAG) to automate financial report summarization, collapsing a multi-hour manual task into minutes. When feature engineering needed to go further, I pushed model accuracy up by 18%. On the infrastructure side, I orchestrate end-to-end data pipelines using Apache PySpark and Apache Airflow, ensuring models retrain on schedule, data flows without failure, and dashboards reflect real-time profit/loss and trade signal performance. I've delivered these in Agile sprints alongside cross-functional teams — contributing to code reviews, scalable ML deployment strategies, and a culture where data drives every decision. What makes my background unusual is the combination: deep statistical modeling, production-grade data engineering, and hands-on Generative AI implementation — all applied to a domain where model accuracy has direct financial consequences. I don't just build models; I build the pipelines that keep them alive. Based in Bangalore and open to full-time Data Scientist roles with teams building serious ML systems — particularly in fintech, AI, or data-intensive product companies. Feel free to message me directly.

Location

Bengaluru

AI stack

antigravity

cursor

langgraph

ollama

Work experience

Data Scientist at NUANCE, Dec 2024 - Present · 1 yr 7 mos, Bengaluru, 1. Designed time-series forecasting models (LSTM, ARIMA, SARIMAX) in production, improving trading trend prediction accuracy and enabling leadership to make more confident, data-driven investment decisions. 2. Automated end-to-end ETL pipelines using Apache PySpark and Apache Airflow to ingest and clean millions of unstructured trading log entries, reducing manual analysis time by 40% and accelerating model retraining cycles. 3. Pioneered a Generative AI solution combining Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG) to automate financial report summarization, compressing a multi-hour manual process to minutes. 4. Engineered advanced feature extraction techniques that boosted predictive model accuracy by 18%, directly enhancing trade signal classification performance and driving measurable improvements in strategy outcomes.

Education

Bachelor of Technology, Electrical and Electronics Engineering, Techno Main - Salt Lake