Backend-focused software engineer with experience building scalable microservices, AI-powered systems, and real-time applications. Experienced in full-stack development, computer vision pipelines, embedded AI systems, and cloud deployment workflows.
Currently focused on:
- Scalable backend architectures
- AI/ML systems with real-world deployment
- Computer Vision using YOLOv8 & OpenCV
- Embedded systems with Jetson Nano & Raspberry Pi
- Flutter-based cross-platform applications
📍 Sri City, India | Sep 2025 – Mar 2026
- Re-engineered a monolithic Node.js backend into a TypeScript microservices architecture
- Reduced average request latency by 90%+
- Improved throughput by 3× under identical load conditions
- Performed scalability validation using load testing
- Worked on production deployment and infrastructure optimization
TypeScript Node.js MongoDB Docker Docker Compose Nginx Cloudinary Twilio GCP
📍 Sri City, India | May 2025 – Jul 2025
- Designed and deployed backend services for:
- Ride booking
- Authentication systems
- Real-time road-quality assessment
- Integrated YOLOv8-based Road Metrics detection outputs
- Built an ML-powered road quality prediction pipeline achieving 93.6% accuracy
- Enabled end-to-end real-time field assessment workflows
Node.js Express.js MongoDB Firebase FCM Twilio Cloudinary Render Python YOLOv8 OpenCV Jetson Nano
🔗 Repository: Autonomous UAV Navigation
- Designed a unified spatio-visual reinforcement learning framework for autonomous map-less UAV navigation
- Developed a multi-scale U-Net perception network for obstacle-aware navigation using monocular RGB input
- Compressed 256×256 visual observations into a compact 28-dimensional navigation state using a 5×5 spatial occupancy abstraction layer
- Implemented and compared PPO and TRPO reinforcement learning agents for continuous UAV control
- Enhanced trajectory smoothness through exponential velocity smoothing and jerk regularization
- Validated the framework in AirSimNH suburban simulation environments with extensive trajectory, reward, and stability analysis
Python PyTorch U-Net PPO TRPO Reinforcement Learning Computer Vision AirSimNH OpenCV Stable-Baselines3
🔗 Repository: EEG Alzheimer's Detection
- Developed an end-to-end FPGA-based machine learning system for Alzheimer's disease detection using EEG signals
- Processed EEGLAB EEG datasets and extracted statistical and frequency-domain features, including Mean, Variance, Alpha, Beta, and Theta Band Power
- Trained a Random Forest classifier and converted floating-point model parameters into FPGA-compatible fixed-point representations
- Manually implemented decision trees in Verilog HDL and designed majority voting logic for hardware-based inference
- Synthesized and deployed the complete architecture on a Xilinx Zynq Edge FPGA board using Vivado
- Integrated LEDs, buzzer, switches, and a 16x2 LCD display for real-time prediction and classification visualization
Python Verilog HDL Vivado Xilinx Zynq FPGA Scikit-learn Random Forest MNE NumPy SciPy EEG Signal Processing
🔗 Live: https://agrivision-livid.vercel.app/
🔗 Repository: AgriVision
- Developed a multilingual Flutter + Flask prototype during a 48-hour hackathon
- Integrated real-time weather and crop intelligence systems
- Built CNN-based plant disease detection with 80%+ accuracy
- Designed scalable API architecture for future production deployment
Flutter Typescript HuggingFace CNN Python Render Dart Node.js Flask TensorFlow Lite Machine Learning REST API MongoDB Twilio JWT Authentication HTML CSS JavaScript
🔗 npm: https://www.npmjs.com/package/logsave-hub/
🔗 Repository: logsave-hub
- Published a Node.js logging and monitoring package with structured logging and real-time observability features
- Built a live web dashboard powered by Socket.IO for log streaming, filtering, search, and historical log analysis
- Implemented automated log retention, file management, and seamless Express integration for production environments
- Designed a modular TypeScript architecture focused on scalability, performance, and developer experience
TypeScript Node.js Express.js Socket.IO Winston npm WebSockets Logging Systems Monitoring Dashboard File Handling
- Computer Vision
- Deep Learning
- Reinforcement Learning
- Embedded AI Systems
- Scalable Backend Infrastructure
- Real-Time Systems
- Microservices Architecture
- Edge AI Deployment
- 🥈 ElectroForge - IoT (Utkrista) — Runner Up
- 🌾 AgriVision (Abhisarga AGRIAI) — Finalist
- 💻 Bit-N-Build Hackathon — Participant
- 📜 Certificate Program 2.0 — IoT & Digital Twin
