I'm a data scientist passionate about building and launching real-world AI products. With a strong foundation in Generative AI, machine learning, and product management, I specialize in translating complex data insights into business strategy and product-led growth.
I build robust, end-to-end data pipelines using the AWS ecosystem (Glue, S3, Bedrock). I leverage PostgreSQL for structured data storage, Docker for containerizing applications, and Git for version control.
Python is my core language. I'm an expert in its data science stack, using Pandas for data manipulation, Scikit-learn for classical machine learning, and frameworks like TensorFlow for building deep learning models.
I bridge tech and business using Jira for agile project management and workflow automation, New Relic and Power BI to create impactful analytics dashboards. I also use FastAPI for API deployment and Notion for documentation.
Analyzed synthetic A/B test data for a hypothetical checkout button change. Performed statistical significance testing and formulated a data-driven "ship / no-ship" recommendation, demonstrating product sense and analytical rigor.
View on GitHub
Developed and deployed a machine learning model to predict customer churn. The process included data loading, cleaning, feature engineering, model training (Logistic Regression), and deployment as a REST API using FastAPI. The API provides real-time predictions and churn probability.
View on GitHub Live API Docs
Built an interactive web application allowing users to upload PDF documents and ask questions about their content. Implemented a Retrieval-Augmented Generation (RAG) pipeline using LangChain, local embeddings, and a ChromaDB vector store to retrieve relevant document sections.
View on GitHub Live App Demo
Architected end-to-end agentic AI platform that transforms problem statements into complete project workflows with strategic task breakdowns, technical documentation, and fully-structured Jira tickets. Directly creates stories, epics, and sub-tasks with contextualized comments, eliminating manual setup and accelerating sprint planning by 40%.
Led development of enterprise AIOps system integrating AWS Bedrock LLMs with Jira workflows to autonomously process 500+ daily support tickets. Engineered ML pipeline: data extraction → preprocessing with AWS Glue → LLM analysis via Bedrock & SAM CLI → API validation → automated Jira updates.
Developed intelligent system leveraging NLP, vector embeddings, and semantic search to surface relevant historical tickets. ML algorithms analyze 50,000+ support cases to match queries with proven solutions, reducing resolution time by 30% and improving first-contact resolution rates.
Developed LLM-powered chatbot serving internal teams and external customers. Architected RAG infrastructure using AWS Bedrock, API Gateway, Glue, Athena, and S3 to deliver real-time, context-aware responses grounded in enterprise knowledge
Engineered Power BI dashboard with live Jira API integration tracking 50,000+ annual tickets and KPIs. Designed visualizations for year-over-year trends in resolution times, SLA compliance, and multi-tier performance.
