Projects

ML & AI Projects

• Credit Risk Analysis: Applied XGBoost and neural networks on Amex data (F1: 93.5%).
• Airbnb Pricing Optimization: Used XGBoost + SHAP for price driver insights in San Diego.
• Algorithmic Trading: Built real-time ETL + LSTM models for price prediction from REST API.

Generative AI & NLP

• Tweet Emotion Detection: LSTM + Hugging Face Trainer to classify tweet sentiments.
• LLM Fine-tuning: DistilBERT, FLAN-T5, Mistral, Gemma using LoRA, QLoRA, IA3 with PEFT.
• BART summarization with semantic search via Azure VectorDB + MiniLM embeddings.

Cloud & Smart Systems

• Cloud-Based Smart Energy: AWS Lambda, DynamoDB for EV charging optimization.
• Power BI dashboards: For manufacturing downtime reduction (Viatris MLOps team).

Embedded & IoT Projects

• Quadratic Buck Converter: Designed & built a prototype; hardware-validated in 2019.
• Solar MPPT: Implemented with 8051 microcontroller during MTech studies.
• 3D Object Scanner: Image processing using MATLAB and OpenCV.

Academic Highlights

• Adobe Analytics Challenge: Collaborated with GM on budget analysis and ad ROI.
• UTA Data Competition Finalist: Optimized ad spend via multi-variable regression.
• Alteryx Datathon: CbyC tax data trends analysis for The Walt Disney Company.