A Deep Learning System for Detecting and Categorizing Anomalies
Log-based anomaly detection for distributed systems. BERT + BiLSTM with attention across HDFS-style logs. Fault isolation and low-latency inference for backend reliability.
Building systems that must respond
before hesitation exists.
Log-based anomaly detection for distributed systems. BERT + BiLSTM with attention across HDFS-style logs. Fault isolation and low-latency inference for backend reliability.
Sensor-driven monitoring architecture for automated fluid tracking and air embolism detection with safety-triggered alert mechanisms.
Hybrid machine learning models for clinical diagnostic classification using multi-parameter medical datasets with ensemble methodology.
Remote robotics platform — browser-based code execution controlling physical robots at IIT Bombay. Not generic web dev.
Industrial robotic fleet monitoring — warehouse-scale autonomous robot supervision. High-frequency telemetry. Industrial software, not dashboard cloning.