Dual Degree Student pursuing excellence in both Computer Science and Electrical Engineering:
- 🎯 BS in Data Science - IIT Madras (Online Degree Program)
- ⚡ B.Tech in Electrical and Electronics Engineering - Amrita Vishwa Vidyapeetham, Coimbatore
- 📚 Certified in Modern Computer Vision & Digital IC Design (SWAYAM)
Hardware acceleration researcher and full-stack developer specializing in AI-driven systems and VLSI design. Passionate about building energy-efficient ML accelerators and deploying production-grade intelligent systems for real-world impact.
- 🔬 Researching Binary Neural Network (BNN) accelerators achieving 6.4 GOPS at 88mW power
- 🏥 Developing ML systems for medical diagnostics (anemia classification, EMG processing)
- 🔋 Building AI-powered Battery Management Systems with intelligent SoC estimation
- 🌊 Creating full-stack data platforms for scientific research and government initiatives
- 🏆 Active hackathon competitor (UIDAI, Smart India Hackathon, startup pitch competitions)
- 📍 Based in Coimbatore, Tamil Nadu, India
Energy-efficient neural network accelerator with custom XNOR-popcount architecture
- Designed systolic array-based accelerator replacing traditional MAC units with XNOR+popcount operations
- Achieved 6.4 GOPS throughput at 88mW power consumption - optimized for edge deployment
- Implemented in SystemVerilog/Verilog with complete RTL design and verification
- Features: Sparsity exploitation, memory hierarchy optimization, pipelined datapath
- Impact: 10x power reduction compared to conventional architectures
- Tech Stack: SystemVerilog, Verilog, Vivado, Python (model training), Hardware-Software Co-Design
Real-time medical diagnostics with quantized deep learning
- Evolved from ensemble methods to quantized deep learning for 5x faster inference
- Deployed on resource-constrained medical devices for point-of-care diagnostics
- Achieved clinical-grade accuracy while reducing model size by 75%
- Tech Stack: Python, PyTorch, Model Quantization, Edge Deployment
AI-driven SoC estimation and active cell balancing for Li-ion packs
- Developed for startup pitch competition - presented to industry judges
- Implemented ML-based State of Charge (SoC) estimation with 98%+ accuracy
- Designed active cell balancing control logic for extended battery lifespan
- Real-time monitoring and safety control for lithium-ion battery packs
- Tech Stack: Python, Control Systems, Embedded Systems, ML Regression Models
Large-scale data analytics for India's biometric database system
- Government of India initiative - January 2026 hackathon participant
- Built specialized modules for processing anonymized Aadhaar datasets
- Developed analytics pipeline for demographic insights and security enhancement
- Handled large-scale data processing with privacy-preserving techniques
- Tech Stack: Python, Pandas, Data Processing Pipelines, Statistical Analysis
Automated fish age estimation using advanced computer vision
- Implemented U-Net with ResNet-50 backbone for precise otolith segmentation
- Automates biological research workflow for marine environmental studies
- Achieved 95%+ segmentation accuracy on complex microscopy images
- Tech Stack: Python, PyTorch, U-Net, ResNet-50, OpenCV, Image Segmentation
Full-stack platform with microservices architecture
- Built Go backend with PostgreSQL database and React.js frontend
- Deployed using Docker and Coolify with separate Python ML service
- Implemented data validation, preprocessing, and analytics dashboards
- Designed for heterogeneous marine biodiversity data ingestion
- Tech Stack: Go, PostgreSQL, React, TypeScript, Python, FastAPI, Docker, Microservices
Real-time muscular signal processing for prosthetic control
- Designed custom Analog Front-End (AFE) for µV-level signal capture
- Implemented real-time signal processing for gesture recognition
- Hardware-software integration for prosthetic limb control applications
- Hackathon project - completed in 48 hours
- Tech Stack: Analog Circuit Design, Signal Processing, Real-time Systems, Python
Rapid prototyping for national-level problem statements
- Competed in India's largest hackathon (December 2025)
- Developed end-to-end solution with hardware-software integration
- Worked in cross-functional team under time constraints
- Tech Stack: Full-Stack Development, Rapid Prototyping, Team Collaboration
- 📝 Publishing research on novel BNN accelerator architectures for energy-efficient AI
- 🔧 Exploring PUF-based security mechanisms for hardware systems
- 🩺 Building production ML pipelines for medical diagnostics and anomaly detection
- 🎓 Completing SWAYAM certifications in Modern Computer Vision and Digital IC Design
- 🚀 Contributing to open-source hardware design tools and ML frameworks
- 🥇 UIDAI Data Hackathon Participant - Government of India (January 2026)
- 🥈 Smart India Hackathon Competitor - National Level (December 2025)
- 🎖️ Startup Pitch Competition - BMS Project Presentation
- 📜 SWAYAM Certified - Modern Computer Vision & Digital IC Design
- 🎯 Multiple Hackathon Finalist - Hardware, Software, and AI tracks
- Energy-efficient AI hardware accelerators
- Binary and quantized neural networks
- Hardware-software co-design for ML systems
- Medical AI and diagnostic systems
- Embedded ML for edge devices
- VLSI design for specialized computing
I'm always open to collaboration on hardware acceleration, ML systems, or interesting hackathon projects!
- 💼 LinkedIn: Vakkalagadda Tanush Pavan
- 📧 Email: tanushpavan2006@gmail.com
- 🐙 GitHub: @HUNT-001