About
Final-year Computer Science and Engineering student with a minor in Business Analytics, building AI/ML applications that connect model-driven decisions with usable software systems. Hands-on project experience in real-time ML risk scoring, TensorFlow-based federated learning, marketplace intelligence, and LLM-assisted asynchronous data processing. Experienced with Python, TensorFlow, PyTorch, FastAPI, Gemini, LangChain, SQL, and Docker, with an interest in developing reliable, deployable AI systems for real-world operational problems.
Projects4
Sentinel: Real-Time Risk Scoring Engine
Built a real-time risk-scoring system combining user, device, and application signals for continuous trust evaluation across 50+ simulated entities with ML-driven threat assessment delivering dynamic policies such as ALLOW, CHALLENGE, RESTRICT, and BLOCK. Integrated live event streaming and Prometheus/Grafana monitoring for scoring activity and security events.
- TensorFlow
- FastAPI
- React
- Supabase
- WebSockets
- Prometheus
- Grafana
Transaction Processing Pipeline
Developed an asynchronous pipeline processing noisy financial CSV files with background workers, implementing data cleaning, statistical outlier detection, currency inconsistency checks, Gemini-based classification, and reporting. Designed persistent job-status and results APIs with PostgreSQL and retry handling to ensure resilience against LLM failures for batches of 5,000+ records.
- Python
- FastAPI
- Celery
- Redis
- PostgreSQL
- Gemini
- Docker
PriceSentinel: AI-Powered Marketplace Intelligence
Built a marketplace intelligence platform tracking competitor prices, Buy Box ownership, price drops, and stockout opportunities across 12+ Indian cities. Integrated price-monitoring pipeline with PyTorch-based Buy Box win-probability prediction and LangChain-powered alert workflow, enabling actionable WhatsApp alerts via scheduled scraping, persistent price data, and REST APIs.
- Python
- PyTorch
- scikit-learn
- FastAPI
- Scrapy
- LangChain
Federated Learning with On-Chain Aggregation
Developed a federated-learning platform featuring neural networks, linear and logistic regression, decision-tree models with federated weight aggregation, model evaluation metrics (precision, recall, F1-score), and hyperparameter search. Integrated blockchain-based aggregation records and IPFS model storage for secure, decentralized coordination.
- React
- TensorFlow.js
- Blockchain
- IPFS
- Firebase
Education1
B.Tech
2023 — 2027KIT’s College of Engineering, Kolhapur, Maharashtra
Computer Science and Engineering