Sagnik Roy

• Migrated reference data setups to highly scalable and dev-friendly backend.
• WIP with symbol mapping across multiple data vendors to standardize classification and temporal analysis.
• Designed and leveraged database-use cases to cater for many queries simultaneously.
• Created seamless historical data regeneration setup to add/correct updates on datasets through the entire historical range. Op

• Primary Tech-Stack: Python, SQL, Bash Scripting, JIRA, DevOps.
• Actively managed reference data crucial for traders during live trading sessions.
Oversaw a comprehensive product spanning multiple (30+) stock exchanges, requiring deep domain expertise.
• Maintained JIRA hygiene in planning while executing tasks.
• Conducted detailed historical analyses to ensure data integrity and reliability.
•

Worked as Data Engineer Intern at AlphaGrep.

• Primary Work-Stack: Python, SQL, GCP, PHP, Confluence
• Established and maintained databases while orchestrating Google Cloud Platform (GCP) workloads via REST APIs, ensuring seamless integration and efficient performance.
• Executed projects based on the Waterfall model methodology, meticulously handling GCP processes, including provisioning, deployment, and optimization.
• Spearheaded the migr

• Primary Tech-Stack: Python, Bash, Machine Learning, async programming
• Engaged in a research and development position involving the review and implementation of cutting-edge models derived from state-of-the-art (SOTA) research papers.
• Primarily focused on computer vision applications utilizing deep learning techniques.
• Developed models trained on bespoke datasets, incorporating transfer lea

• Primary Tech-Stack: Python, C++, Web-Crawlers and Scrapers, Google Collab
• Utilized web scraping techniques to gather freely available data, subsequently enhancing dataset sizes through various methods such as data augmentation, duplication using common image processing techniques, and generative deep learning approaches.
• Generated random tabular datasets from scratch for HackerEarth challeng

• Had been into insightful sessions given by industry builder deep learning and machine learning geniuses.
• Learnt about basics and advanced of Machine Learning from gradient updates to reinforcement learning

• Primary Tech-Stack: Python, SQL, Bash Scripting, HDFS, Jupyter Notebook
• Managed cron jobs for scheduled tasks, ensuring the timely execution of essential processes.
• Developed efficient data processing pipelines using Spark, pandas, and bash scripting, optimizing workflows for enhanced performance.
• Successfully engineered cross-product recommendation systems, delivering personalized recomme

• Primary Tech-Stack: Python, CVAT, LabelImg, MakeSense, Google Collab
• Annotated over 30 datasets for tasks such as instance segmentation, object detection, and key point detection, spanning various domains from government-related projects to sports analytics and warehouse management.
• Utilized efficient annotation tools such as LabelImg, CVAT, and LabelMe to expedite the annotation process.
•

• Searched for real-life problems that can be fully or partially optimized by Machine Learning.
• Developed tutorial notebooks on the same problems to showcase applied Machine Learning.