Data Scientist

Overview

Data Scientist for the AI & ML (Data Collection) team, building AI-powered solutions to extract structured information from PitchBook reports, news, and other content.

What you'll do

• Own end-to-end data science initiatives from problem definition and data exploration through monitoring and continuous improvement.
• Translate business requirements into measurable data science problems and define success criteria.
• Develop NLP, machine learning, and LLM solutions for document understanding and information extraction.
• Build extraction workflows using document parsing, embeddings, RAG, prompt engineering, fine-tuning, and agentic approaches.
• Create evaluation datasets and metrics (e.g., precision, recall, F1, field-level accuracy, coverage, confidence) and conduct error analysis.
• Partner with ML engineers to integrate solutions into production and monitor quality, regressions, and drift.
• Collaborate with product, engineering, platform, and domain teams to deliver scalable solutions.

What you'll need

• 2+ years of experience in applied data science, machine learning, NLP, or information extraction.
• Bachelor's or Master's degree in a quantitative field such as Data Science, Computer Science, Statistics, Mathematics, Economics, Engineering, or related.
• Experience analyzing large complex structured and unstructured datasets including exploration, preprocessing, feature engineering, sampling, labeling, and dataset construction.
• Hands-on experience building document intelligence or information-extraction solutions using transformers/embeddings, RAG, LLMs, prompt engineering, fine-tuning, or agentic workflows.
• Understanding of experimental design, statistical reasoning, model evaluation, and error analysis using metrics like precision/recall/F1.
• Proficiency in Python and SQL, with experience using pandas, NumPy, scikit-learn, and PyTorch or TensorFlow.
• Experience with Hugging Face, LangChain, or comparable NLP/LLM frameworks.

Details

• Location: Mumbai.
• Work setting: standard office; hybrid work model with in-person collaboration each week.