Raghav Srivastava
**.*********@*****.*** — +91-981******* — New Delhi — LinkedIn — GitHub
Education
IILM College of Engineering & Technology Greater Noida, India B.Tech in Computer Science & Engineering 2021 - 2025 — CGPA: 8 Experience
Tara Capital Partners — Quant Researcher Delhi — Oct 2023 - Present
• Deployed and currently managing live books for mid-frequency trading based on machine learning in Indian equity futures and event-driven, cross-sectional momentum strategies in Indian and US markets, achieving 3+ Sharpe ratios.
• Developed and backtested trading strategies for commodities futures across 17 exchanges, achieving 1.5+ Sharpe ratios with
<15% max drawdowns.
• Built low-latency pipelines for live and historical order book data storage and processing with sub-50 ms latency.
• Developed software infrastructure for backtesting and paper trading live intraday strategies in the US and Indian markets. StockBrain — Founder Remote — Sept 2023 - Oct 2023
• Leveraged deep learning to model price variations in NIFTY50 stocks.
• Developed technical indicator-based strategies for MFT in equities, arbitrage strategies for 7 major forex pairs. Nethermind — Data Engineering Intern Remote — July 2023 - Sep 2023
• Optimized data pipelines, increasing high-frequency transaction processing speed by 5%.
• Created a Python library for Starknet contract data encoding/decoding, enhancing transaction indexing efficiency by 10%. AlgoAnalytics — Quant Intern Remote — May 2023 - July 2023
• Integrated sentiment scores from FinBERT, ROBERTA, BERT based on Twitter data for NIFTY50 stocks into an XGBoost model, boosting accuracy by 2% & strategy returns by 0.3%. Pace Stock Broking Services Pvt. Ltd. — Quant Intern Delhi — Dec 2022 - Feb 2023
• Developed a backtesting framework, reducing strategy bactesting time by 50%.
• Automated access token retrieval for Zerodha & XTS, enabled real-time tick data processing via websockets with a latency reduction of 30 ms.
Georgia Institute of Technology — Research Intern Remote
• Developed an Economic Entity Recognition (EER) model for financial documents, increasing accuracy by 3%.
• Benchmarked & fine-tuned LLMs (LLaMA-2, GPT-3.5), improving financial data processing efficiency by 10%. IIT Jodhpur — Research Intern Remote
• Advanced reinforcement learning with a cognitive model for curiosity, improving simulation accuracy by 8%. Skills
• Programming: Python, SQL, C++, C#, MATLAB
• Libraries/Frameworks: Scikit-Learn, TensorFlow, Keras, NumPy, Pandas, Matplotlib, PyTorch, Scipy, NLTK, Dask
• Tools/Platforms: Git, Linux, AWS, Docker, Hugging Face, Kubernetes, Langchain, Flask, Streamlit, MS Excel, Bloomberg, CapitalIQ, Refinitiv, Polygon
• Databases: PostgreSQL, SQLite, MongoDB, ChromaDB Projects
MorphAI — Link
• Developed LLM-based chatbots with document retrieval and web searches, reducing response time by 15%.
• Fine-tuned and integrated RAG pipeline with GPT-4, GPT-3.5 T, Llama2, etc.
• Achieved 2X faster research & analysis, 4X productivity, 30% cost savings, 98% accuracy, 4.8/5 customer satisfaction rating. Black-Scholes Option Pricing Model — Link
• Streamlit app implementing the Black-Scholes option pricing model.
• Calculated and visualized option prices and their sensitivities based on input parameters. Bank Fraud Detection — Link
• Implemented a logistic regression model for bank fraud detection to flag illegal transaction attempts, trained on 3.2 million data points, achieving 99.9% accuracy.
Certifications
• Advanced SQL - Kaggle
• Text Classification Model with AWS Glue and Amazon Sage- Maker - AWS
• Advanced Software Engineering Virtual Program - Walmart Global Tech
Honors & Awards
• Google Developer Student Club Lead at IILM University (2023-24)
• First rank in national-level Vedic Maths competition