Building intelligent systems at the intersection of machine learning, data engineering, and backend development. Currently working at Klimashift and pursuing M.Tech at IIIT Gwalior.
I'm a ML Engineer & Data Scientist with a B.Tech in CSE (Data Science & ML) from ITM University and currently pursuing M.Tech at IIIT Gwalior with a CGPA of 8.21.
I specialize in building end-to-end ML pipelines, computer vision systems, and LLM applications. At Klimashift, I engineer energy disaggregation models and automated asset identification pipelines that run in production on AWS.
Beyond ML, I work across the full stack — from Flask/Django/FastAPI backends and PostgreSQL databases to cloud infrastructure on AWS and Azure. Good engineering means solving the right problem, cleanly.
Non-Intrusive Load Monitoring ML model for energy disaggregation with ETL pipelines processing large-scale time-series datasets on AWS. Identifies appliance-level consumption from aggregate meter readings.
Automated computer vision pipeline using YOLO, Grounding DINO, and Grounded-SAM to detect, segment, and catalogue commercial assets from raw image frames — eliminating manual inventory effort entirely.
MLflow-tracked pipeline with automated performance dashboards, data drift detection, and CI/CD via GitHub Actions — surfaces model degradation before it impacts production.
Flask-based facial recognition web app achieving 92% accuracy with OpenCV deep learning. REST API integration for real-time data storage, user authentication, and audit-trail-ready PostgreSQL backend.
End-to-end ML pipeline comparing SVM, Logistic Regression, and Decision Tree classifiers on financial data with feature engineering. Achieved 94% classification accuracy. Full evaluation with ROC-AUC and F1 metrics.
Analyzed end-to-end supply chain data in Celonis EMS with trend deduction and pattern recognition via ETL and REST APIs. Reduced business decision cycles from 4–5 days to 2 days.
ML system to detect and classify DoS attacks in Wireless Sensor Networks using anomaly detection and supervised classification on network telemetry features.
Java GUI application integrated with MySQL with optimised database queries reducing end-to-end transaction time by 50%. Includes inventory management, billing, and reporting modules.
I'm open to full-time roles, research collaborations, and freelance ML/data engineering work. Whether you have a project in mind or just want to talk tech — reach out anytime.
I typically respond within 24 hours.