Sanjay Sathyapriyan

About Me

Motivated AI Engineer with over 4 years of experience across Generative AI, machine learning, and software development. Holds a Master’s in Business Analytics from UT Dallas and certification in AWS Cloud. Skilled in LLMs, RAG, Python, LangChain, and AWS, with hands-on experience building NLP pipelines, ML models, and GenAI-driven tools in finance, real estate, and enterprise software.

Current location: Edison, New Jersey

Preferred location: Open to Relocate anywhere within USA

Work Experience

Seeking AI Engineer, Data Scientist, Data Analyst, or Business Intelligence Engineer roles

Professional Experience

AI Engineer

Feb 2025 - Present

Illuminate AI ---- Remote, USA

Developed an end-to-end asynchronous crawl4AI Python pipeline that crawls 25+ AI news sources, uses LLMs for content extraction and impact-driven summarization, reducing review time by 90% while maintaining a <5% duplication rate.
Optimized deployment by integrating the pipeline into a scheduled GitHub Actions workflow and replacing DeepSeek with lighter LLMs cutting per-article processing cost by 33% (from $0.30 to $0.20) through more efficient deduplication and summarization.
Engineered a Python-based content generation pipeline to auto-create LinkedIn post templates and Instagram reel scripts from AI news updates, saving 10+ manual hours weekly and accelerating content delivery across platforms.
Architected a full-stack web application using Node.js and JavaScript to capture subscriber data and trigger automated, personalized emails scaling outreach to 100+ recipients daily and improving user engagement through tailored messaging.

Business Data Analyst

OCt 2024 - Nov 2024

UnitedHealth Group (Contract) ---- Remote, USA

Conducted detailed data analysis and stakeholder interviews to map AS-IS/TO-BE process flows using Visio uncovering 10+ automation opportunities that resulted in a weekly reduction of 12+ hours of manual effort
Translated business needs into data requirements and created functional specifications to support IT and analytics teams, ensuring alignment between operational goals and data-driven initiatives.

Data Analyst (GenAI)

May 2023 - May 2024

Trepp Inc. ---- Irving,TX

Compared costs and benefits of LLMs (GPT, LlaMa, Claude) adoption for real estate use cases, presenting findings that led executives to approve three new GenAI features
Built a customer support chatbot POC using LangChain and gpt-3.5 demonstrating potential for automating responses to common questions and routed cases to correct teams, potentially reducing support workload by 15 hours weekly
Assisted in testing early RAG approaches against finetuning methods for adding context to user queries, documenting findings that suggested RAG could improve answer relevance while requiring less computational resources
Developed a regression-based machine learning model in Python to predict property financials backed by physical attributes and local market dynamics, achieving 86% accuracy in projecting financials.
Designed Tableau dashboards visualizing CRE market trends across 5+ cities and 4 property types, increasing average client satisfaction score by 15% measured through feedback surveys

Software Engineer

Dec 2019 - Jul 2022

Infosys Limited ---- Chennai,India

Modernized a client fleet maintenance web application using Angular 8, HTML, and CSS, reducing UI latency by 15% and improving overall user workflow efficiency
Fixed Java backend performance bugs by refactoring service components and adding unit test coverage with Mockito and JUnit, improving application response times by 20% and solving 5+ open customer tickets
Created Python script with MySQL queries to track slow-moving inventory across 8 warehouses, setting up weekly alerts that reduced dead stock by 15%.
Established standardized project documentation using Confluence, cutting employee onboarding time by 40%.

Analytics Projects

AI-Driven Knowledge Retrieval and QnA Chatbot - LLM, RAG, Streamlit, Pytorch

Developed an AI-driven chatbot for text generation, summarization, and QnA by utilizing foundation models (FM)
Used Amazon Bedrock APIs and LangChain to develop applications while implementing Retrieval-Augmented Generation (RAG)
Built a frontend interface with Streamlit to converse with the Large Language Model (LLM) and presented it to the AWS club while engaging over 40 participants, addressing seven in-depth queries

Credit Risk Prediction - Machine learning, Python, XGBoost, Neural networks, FMCG

Conducted data cleaning and feature engineering, utilizing feature importance scores from XGBoost to develop custom models
Utilized grid search for parameter optimization, achieving a 0.8944 AUC score on test data for the XGBoost model
Inferred that the Repayment variable was the most important feature for predicting default, and the XGBoost performed better than the Neural Networks

Conagra Brands Market Analysis - Statistics, OLS Regression, Python, R, Tableau

Led a team of 6 to analyze market-level sales data and customer reviews of competitor brands in the FMCG table spreads category to identify potential growth areas via merchandising
Implemented an OLS regression model to quantify interaction effects between sales and merchandising of comparable competitors, providing actionable recommendations to Conagra for product merchandising

Employee Attrition HR Analytics - MySQL, MongoDB, ERD

Led a team of 5 to design and normalize a 3NF relational data model for analyzing employee attrition with MySQL and MongoDB
Designed database table schemas and formulated an entity relationship diagram (ERD) to make business sense

Certifications