Sayak Banerjee
Quantitative Researcher | NLP/ML @Carnegie Mellon, SCS
- Role
- Data Scientist at Paylocity
- Location
- New York, NY, US
- LinkedIn followers
- 500 followers
About Sayak Banerjee
Grad student of the Master of Computational Data Science (MCDS) program at the School of Computer Science, Carnegie Mellon University with a specialization in the Analytics Track. My academic and professional pursuits are driven by a deep interest in Data Science, Machine Learning, Natural Language Processing, and AI. My research interests are focussed in language models and information retrieval.Over the summer I interned at Paylocity as a Data Scientist - where I am built out a RAG based AI assistant and actively worked on Search Optimization.With over two years of experience in handling alternative financial data for a US-based Hedge Fund, I specialize in data science and data engineering. My technical expertise includes Python, Tensorflow, C++, PySpark, SQL, Snowflake, Machine Learning and Natural Language Processing.In addition to my professional experience, I have a strong background in research within the fields of data science and machine learning. I have published five research papers in Scopus-indexed journals and have presented my work at conferences.
Experience
Data Scientist
May 2025 — Present · Pittsburgh, PA, US
Collaborating with the Search team to enhance an enterprise chatbot solution utilizing RAG architecture focussing on queries related to the Paylocity portal, IRS-compliance and company handbooks.• Improved the existing re-ranking pipeline by removing BM25, which resulted in an average reduction of 5k tokens across user queries and reduced latency by ~2 seconds. Further, enhanced the initial ranking pipeline by incorporating and finetuning multi-match keyword-based search alongside HNSW ANN search, creating a hybrid ranking pipeline.• Benchmarked Azure OpenAI embedding models against Amazon Titan v2 (Bedrock) and offline models from HuggingFace to evaluate performance and latency. Implemented a dynamic fallback mechanism to switch to Bedrock embeddings during OpenAI latency spikes or outages, improving the AI assistant’s SLA by ~2% for our customers. This work laid the foundation for upcoming load balancing and model routing enhancements.
Education
South Point High School, Kolkata
Mathematics and Computer Science
Carnegie Mellon University
Master of Computational Data Science
Vellore Institute of Technology
BTech - Bachelor of Technology, Electronics and Communications Engineering
Find verified contacts for anyone on LinkedIn
Unifers gives sales teams verified emails and direct dials, enriched profiles, and outreach that lands in the inbox.
Free plan included · No credit card required
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.