Risk-X: Real-Time Risk Intelligence
Cloud-native streaming system for transaction risk monitoring: online anomaly scoring, drift detection, and automatic threshold/model adaptation for stable alerting.
Cloud-native streaming system for transaction risk monitoring: online anomaly scoring, drift detection, and automatic threshold/model adaptation for stable alerting.
Safety-first humanoid navigation demo: camera frame → VLM decision → one controlled ROS motion primitive. Conservative defaults, predictable behavior, easy debugging.
On-device pose estimation + interpretable attention heuristics running in real time on a Raspberry Pi. Logged 10k+ frames for future learning-based upgrades.
Co-founded a drone battery-swapping prototype project: CAD + hardware integration + control systems. Supported by incubators and recognized at Idea Fest events.
Random Forest regression with feature engineering + evaluation (MAE/RMSE/R²) to predict flight prices. Includes feature importance analysis.
Linear regression on e-commerce behavior metrics to predict yearly spending and interpret which features drive purchases.
Decision tree classifier for prescription prediction with preprocessing, cross-validation, hyperparameter testing, and model visualization.
Hands-on data science workflow: Python, SQL, visualization, ML foundations, and a capstone project.
Core Python fundamentals including data structures, algorithms, and object-oriented programming.
Data organization, analysis, and visualization using formulas, pivot tables, and charts.
Intro ML concepts: supervised/unsupervised learning, preprocessing, training, and evaluation workflows.
Recognized for Welkin Aves; awarded a grant of 50,000 INR to support further development.
Awarded a grant of 2,50,000 INR to advance the prototype toward real deployment.
Placed first in a competitive coding / problem-solving event held across engineering campuses in Kerala.