I'm a final year Electronics and Computer Engineering (ECM) student at VIT Chennai. Most of my work sits in Data Science and Machine Learning—training models on tabular and sensor data, engineering features, and building forecasting pipelines. I also focus on data validation, SQL schema design, and writing clean APIs to turn models into working systems.
Programming Languages
 |
||
|---|---|---|
| Java | SQL | Python |
Machine Learning & Deep Learning
 |
|||
|---|---|---|---|
| PyTorch | TensorFlow | Scikit-Learn | Keras |
Architectures & Methods: LSTM • Time-Series Forecasting • Random Forest • CNN • Feature Engineering
Computer Vision
 |
 |
||
|---|---|---|---|
| OpenCV | ROS 2 | MediaPipe | YOLOv8 |
Edge & Real-Time Perception: Sensor Fusion • Monocular Distance Estimation • 6-DOF IMU Telemetry • Spatial Filtering
Databases & Data Engineering
| SQLite | MySQL | PostgreSQL |
Core Pipelines & Backends: REST APIs • Automated ETL Ingestion • Data Validation • In-Memory Caching
Data Tools & Analytics
 |
 |
 |
 |
|---|---|---|---|
| Pandas | NumPy | Matplotlib | Power BI |
Analytics & Decision Modeling: Statistical Modeling • Multi-Criteria Decision Analysis (MCDA) • Time-Series Error Benchmarking
Developer Tools & Environment
| Docker | Linux | Git | GitHub | GitHub Actions | VS Code |
| Project | Core Stack | What It Does & Key Numbers | Links |
|---|---|---|---|
| TRIDENT | SQLite, LSTM, Docker, REST APIs | Ingests 17,500+ hourly weather records via automated ETL. Built a 2-layer stacked LSTM (MSE < 0.005) forecasting 24-hr hazard trends and an MCDA triage engine auto-prioritizing emergencies across 10 zones. | Repo • Live App • Demo |
| SPARC | MediaPipe, Random Forest, CNN, OpenCV | Sign language recognition engine running at 15–20 FPS. Engineered a 229-dimensional geometric feature space from 920k raw visual dimensions, hitting 92.4% accuracy and cutting false positives by 92% across 41 classes. Top 5 National Finalist at Smart India Hackathon 2025. | Repo • Live App • Demo • Certificate |
| stembot | ROS 2, EKF, Python, LiDAR | Sensor fusion and filtering pipeline built during my robotics internship at STEMTEC. Validated a 15 Hz Extended Kalman Filter against raw encoder logs, cutting spatial drift by 60% (0.45m → 0.18m) and angular drift by 69.2%. Filtered ~3,600 LiDAR points/sec with 95% obstacle detection accuracy under 50ms latency. | Repo • Certificate |
| NeuroBridge AI | SQLite, MediaPipe, REST APIs | Telehealth platform with real-time video and multimodal clinical risk scoring (40% vision + 30% behavioral + 30% engagement) using 468 facial landmarks. Includes OLS risk forecasting, TF.js text moderation, and 70+ REST endpoints. | Repo • Live App |
| GNOME Stocks Suite | Python, SQLite, REST APIs | Financial data engine pulling from 3 APIs into a single normalized schema. Built a 3-tier cache (/dev/shm memory + SQLite + TTL) indexing 150k instruments down to 2,500 liquid equities, cutting query latency from 800ms to <5ms with 99.9% uptime. | Repo • Live App |
| DrishtiGuide | YOLOv8, OpenCV, Sensor Fusion | Assistive edge vision pipeline fusing YOLOv8 object detection with monocular distance estimation. Cuts inference overhead by 66.7% at sub-100ms latency, paired with a 6-DOF IMU fall detection setup (>95% reliability). | Repo |
- Top 5 National Finalist, Smart India Hackathon 2025 — SPARC multi-modal perception engine selected out of 500+ teams nationally. [Demo] [Live App] [Certificate]
- 2nd Runner Up (3rd Place), Technovation 2025 — Ranked out of 50+ teams at Kurukshetra ’25, CEG Guindy, Anna University (₹15,000 award). [Certificate]
- Winner, Innovate X Impact (V-NEST) — Secured 1st Position out of 35 teams (₹5,000 award).
- Operations Lead, Team Namo Nirvana (VIT Chennai) — Led operations and execution for our university robotics special team (100+ members) across 5–6 national competitions per semester, boosting operational efficiency by 35% and securing 2 sponsors. [LOA] [Certificate] [LOR]
- Interests & Activities: National & University-Level Basketball player, Ex-NCC Cadet (1 TN BN, Corporal), active social outreach volunteer (VITeach), and occasional quizzer (VIT Chennai Quiz Club).