I build AI systems that reason over data, not just predict.

Hi, I'm Neehanth. As a Data Science professional and AI Enthusiast, I focus on bridging the gap between promising models and reliable, production-ready systems.

View My Work
Read my technical writing:
Medium Substack

About Me

I am a Data Science professional passionate about building intelligent systems that solve real-world problems. My goal is to bridge the gap between promising machine learning models and scalable, production-ready architectures.

I have hands-on experience building LLMs, RAG pipelines, and agentic workflows. I enjoy developing reliable backend services with FastAPI, working with vector databases, and learning to deploy maintainable solutions on cloud infrastructure. I'm eager to bring my curiosity and engineering mindset to a collaborative team where I can continue to grow and deliver impact.

University of Memphis Memphis, TN
M.S. in Data Science Jan 2024 – Dec 2025
Neehanth Reddy

Capabilities

Industry Tech Stack

Core Languages & Frameworks

Python SQL FastAPI REST APIs JWT Authentication pandas NumPy HTML CSS JavaScript

AI, Machine Learning & LLMs

RAG Pipelines Agentic Workflows LangChain LangGraph OpenAI API Ollama Prompt Engineering PyTorch TensorFlow Scikit-learn Hugging Face Transformers Statistical Modeling Computer Vision NLP Feature Engineering EDA

Data Engineering, BI & Databases

PostgreSQL MySQL MongoDB ChromaDB PySpark SQLAlchemy Alembic Power BI Streamlit Data Visualization

Software Engineering & MLOps

AWS Docker GitHub Actions DVC Pytest Pydantic Git GitHub Postman Postman

Experience

Where I've Worked

Data Science Fellow & Graduate Assistant

University of Memphis
Jan 2024 – Dec 2025 Memphis, TN · Hybrid
  • Instruction & Mentorship Served as Graduate Assistant for Advanced Statistical Learning II and Bio-Statistical Learning Methods II. Mentored cohorts in ML using R and SAS; code reviews reduced rework by 20% and lifted exam scores by 12%.
  • Applied AI & LLM Projects Designed a RAG pipeline (LangChain, ChromaDB) for querying 100+ research papers with sub-500ms response times, and an LLM-based agent for natural language database operations with granular permission controls.
Python LangChain ChromaDB RAG FastAPI AWS Docker GitHub Actions R SAS
Leadership & Community
  • Generative AI Specialist Community Dreams Foundation
    Feb 2026 – Present

    Developing AI-driven solutions to streamline community development initiatives and social impact.

  • Organizer FAU Google Developers Group (GDG)

    Leading technical workshops and fostering a local ecosystem for AI and emerging technologies.

Projects

GenAI Research Assistant (RAG Pipeline)

Problem: Researchers struggled to query insights efficiently across hundreds of dense academic papers.
Solution: Designed a RAG architecture utilizing sentence-transformers and ChromaDB to enable precise semantic search.
Highlight: Sub-500ms query response time across 100+ papers using asynchronous FastAPI.
AI / LLMs

AI-Powered Inventory Management

Problem: A multi-warehouse operation needed a reliable way to manage complex inventory rules while allowing natural-language querying.
Solution: Developed a flexible microservice using a hybrid PostgreSQL/MongoDB schema, paired with an intelligent LangChain-based agent.
Highlight: Integrated safe tool-level permissions for the AI agent alongside strict JWT auth.
AI / LLMs

Poultry Disease Classification System

Problem: Early identification of poultry diseases from images was completely manual and prone to high error rates.
Solution: Trained a VGG16 transfer-learning model capable of classifying multiple disease states accurately.
Highlight: Implemented DVC for reproducible versioning and deployed the 94%-accurate model via GitHub Actions to AWS EC2.
Machine Learning

Telecom Customer Churn Prediction

Problem: Extreme class imbalance in customer data caused the business to miss critical at-risk churners.
Solution: Applied SMOTE balancing techniques alongside extensive feature engineering to predict customer churn accurately.
Highlight: Improved churn recall from 0.68 to 0.79 using fine-tuned models without excessive false positives.
Machine Learning

E-Commerce Analytics & Forecasting

Problem: Sales planning was hampered by inaccurate demand forecasts and a lack of granular customer segmentation.
Solution: Analyzed thousands of transactions using clustering and association rule mining, complemented by a SARIMA time series model.
Highlight: Delivered an 18% improvement in sales forecasting accuracy while surfacing actionable customer segments.
Data Science

Scalable Spark Analytics Platform

Problem: Processing over 50 million taxi trip records on standard local hardware was computationally infeasible.
Solution: Provisioned a multi-node Apache Spark cluster to handle distributed data ingestion and analysis.
Highlight: Built resilient PySpark pipelines with built-in schema validation and outlier detection across the 50M+ record dataset.
Data Engineering

Customer Churn Analysis Dashboard

Problem: Needed to uncover churn patterns and quantify revenue impact within a 7,043 telecom customer dataset.
Solution: Developed an interactive Power BI dashboard by integrating Python preprocessing with SQL queries to calculate key KPIs.
Highlight: Pinpointed key churn drivers (month-to-month contracts, senior citizens) and equipped stakeholders with actionable retention insights.
Data Analytics

Writing & Publications

Ideas, Insights & Engineering Deep Dives

From raw data to working systems. Practical stories on ML, intelligent search, AI assistants, and decision tools.

Medium Blog

Simple, practical stories on data science, intelligent search, AI assistants, and bringing ML models into real use.

Read on Medium

Substack Newsletter

Simple breakdowns of ML, AI assistants, and decision tools. Includes curated resources for developers building in this space.

Read on Substack

Want to talk about AI systems, LLMs, or building intelligent products?

neehanth@portfolio ~

❯ neehanth --contact

Email ›

Location › United States

Status › Open to opportunities

❯