Featured Projects
Telecom Customer Churn Prediction
● Built an end-to-end ML pipeline for churn prediction including data cleaning, EDA, feature preparation, class imbalance handling, training, and evaluation.
● Used SMOTE / SMOTEENN and Optuna.
● Deployed with Streamlit on AWS EC2.
Python
Scikit-learn
SMOTE / SMOTEENN
Optuna
Streamlit
AWS EC2
PCAP StoryTeller
● Built a Python-based network analysis solution that converts packet capture data into structured, human-readable troubleshooting narratives.
● Focuses on reducing manual investigation effort for support teams.
Python
Packet Analysis
Data Parsing
Automation
NLP Concepts
AI-Powered Customer Support Copilot
● Built a support copilot using memory, retrieval, and tool-calling concepts.
● Modular FastAPI backend + Streamlit UI + Dockerized workflow.
● Designed for faster contextual support responses.
Python
LangChain
Mem0
FastAPI
Streamlit
Docker