Projects
Selected academic and applied work across vision, ML, and systems.
FinRAG: Retrieval-Augmented Generation for Financial QA
Built a financial-domain RAG system that answers investment, risk-analysis, and forecasting queries with improved factuality and reduced hallucinations.
- Evaluated Mistral-7B, Phi-3, and Llama-3.1 for accuracy, stability, and explanation quality.
- Implemented dense vector retrieval (FAISS), model pipelines, and experiment tracking; deployed on GitHub.
Aug–Dec 2025 • Python, PyTorch, FAISS, LLMs
Repository
Instance Segmentation Using Mask R-CNN (Pen, Plug, Shoe)
Implemented a three-class instance segmentation pipeline on a custom dataset using Detectron2/Mask R-CNN.
- Trained & fine-tuned models; analyzed mAP, IoU, and class-wise performance.
- Prepared results and documentation for public release.
Aug–Dec 2025 • Mask R-CNN, Detectron2, PyTorch
Tango Puzzle Solver with AC-3 and A* Search
End-to-end solver for the Tango logic puzzle with interactive visualization.
- AC-3 for arc-consistency and A* for search; live Pygame view of pruning & backtracking.
- Extended with Q-learning experiments for automated playability.
Feb–May 2025 • Python, Constraint Satisfaction, A*, AC-3, Pygame
Design and Simulation of a Secure Banking Network
Designed a secure multi-branch banking network with IPsec VPN and AWS components to ensure encrypted inter-branch communication.
Sep 2022 • AWS (EC2, VPC, VPN), Packet Tracer, Routing/Switching
Wi-Fi-Based Secured Wireless Communication Using RSA
Implemented RSA-based secure message transmission using Arduino/Atmega, Wi-Fi module, and LCD display.
Oct 2021 • RSA, C, Arduino IDE
