New York City Skyline

Hey,
I am Subiksha

Data Scientist & AI Strategist

Transforming complex data into intelligent solutions. Currently architecting agentic AI systems in the heart of New York City.

Work With Me See Selected Works

About Me

I am a Data Scientist and Pace University Graduate (GPA 3.75) with over four years of professional experience in high-impact data roles. My expertise lies at the intersection of Agentic AI, RAG architectures, and scalable MLOps pipelines.

I have worked on end-to-end data lifecycles for global brands—from optimizing commercial policy discovery at PepsiCo using LLMs to engineering Medallion Architecture pipelines for Levi Strauss & Co. Beyond model development, I am proficient in data visualization and business analysis, ensuring that complex mathematical foundations are translated into clear, actionable insights that drive measurable value.

Education

M.S. Data Science, Pace University

Location

New York, NY

Subiksha Profile

Professional

Experience

Data Scientist | Client: PepsiCo

Cognizant Technology Solutions — Remote/NYC

Oct 2023 – Jul 2024
  • Designed an AWS-based RAG system for commercial policy discovery using LangChain and vector embeddings, reducing manual research latency by 40%.
  • Built robust AWS ML pipelines using Lambda and SageMaker to automate model deployment and monitoring, reducing maintenance overhead by 80%.

Data Scientist | Client: Levi Strauss & Co.

Cognizant Technology Solutions — Chennai, India

Aug 2022 – Sept 2023
  • Engineered ETL pipelines in Databricks using dbt to integrate disparate retail datasets into a Medallion Architecture.
  • Developed customer segmentation and churn models using LightGBM on Spark, achieving 84% recall for targeted retention campaigns.
  • Fine-tuned BERT models for sentiment analysis of customer feedback, delivering actionable insights for seasonal merchandising strategies.

Junior Data Scientist | Client: Nestlé Waters

Cognizant Technology Solutions — Chennai, India

Oct 2020 – Jul 2022
  • Utilized SQL and Tableau/Power BI to surface supply chain KPIs and distribution risks, resulting in over $200k in annual savings.
  • Executed ad-hoc statistical experiments on conversion funnels, communicating data-driven recommendations to cross-functional stakeholders.
  • Developed a modular Python framework for automated data processing and use-case analysis, reducing processing latency by 95%.

Portfolio

SELECTED WORKS