AVAILABLE FOR FREELANCE & ENGINEERING CONTRACTS
ATHARVA THOMBARE
L&T Technology Services | SRM Institute of Science and Technology, B.Tech ECE | Machine Learning Engineer & Full-Stack Developer
L&T Technology Services | SRM Institute of Science and Technology, B.Tech ECE | Machine Learning Engineer & Full-Stack Developer
L&T Technology Services | SRM Institute of Science and Technology, B.Tech ECE | Machine Learning Engineer & Full-Stack Developer
Applied Machine Learning Engineer and Full-Stack Developer with an Electronics & Communication Engineering foundation from SRM Institute of Science and Technology.
Industry experience at L&T Technology Services (Aerospace & Rail Division) and Mettler Toledo India building deep learning models (YOLOv8, Vision Transformers, DenseNet) integrated into production Spring Boot, React, and Flutter applications.
Published IEEE Xplore papers (2026) in applied deep learning & remote sensing.
Extensive empirical benchmark evaluating convolutional neural network topologies against baseline logistic regression for diagnostic MRI brain tumor identification. Demonstrated DenseNet121 feature reuse efficiency while optimizing real-time latency on MobileNetV2.
Multimodal spatial-temporal deep learning framework leveraging Sentinel-2 multispectral imagery fused with meteorological dataset parameters to forecast dengue outbreak risks across urban hotspots using spatial Vision Transformers and 3D CNNs.
Production toolsets across Backend, Mobile Android, Machine Learning, and Embedded Systems.
Multimodal deep learning framework fusing Sentinel-2 satellite imagery with climate data to forecast dengue outbreak risk across 9 Indian cities, with evidential uncertainty quantification and a live GradCAM-driven Streamlit risk map.
Production pipeline combining optical-flow frame interpolation with ESRGAN super-resolution — 2× frame rate, 4× upscaling, 60% faster on 4K via GPU acceleration. PSNR 32.5dB, SSIM 0.92.
5-DOF robotic manipulator equipped with a high-resolution optical sensor and a real-time geometry classification model to sort industrial items on conveyor belts with sub-millimeter precision.
Building a React web application and fine-tuning a YOLOv8n model on custom-annotated data to detect targeted objects on railway tracks. Developing a Flutter application with a Java backend for pulling cloud telemetry into a live monitoring dashboard.
Analyzed industrial communication protocols (IO-Link, PROFIBUS, PROFINET, Ethernet/IP) for checkweigher systems across Siemens and Rockwell PLC architectures. Performed precision load-cell calibration for IND-series terminals and designed SCADA Power BI dashboards.