Avilash S

I’m an AI/ML Engineer specializing in production-grade intelligent systems across NLP, IoT, and agentic automation. I’ve been a finalist in 4 national hackathons (SIH 2024 & 2025, Tamil Nadu Hackathon 2026), and I’m currently pursuing my B.Tech in AI & Data Science (GPA 8.14/10, expected May 2027).

In VoltIQ, I built an EV Battery Intelligence Platform using Python, FastAPI, Streamlit, and scikit-learn, trained on 8M+ records. Our smart modules included SoH prediction (R²=0.9564, MAE=0.10%), thermal anomaly detection on 1M+ records, smart charging optimization, and a CO₂ efficiency tracker, with a brand-agnostic CAN bus normalization layer deployed as a FastAPI backend (6 endpoints) plus a Streamlit dashboard.

In VedaSense, I created an AI-Powered Electronic Tongue with a portable IoT + AI pipeline using ESP32 sensor fusion and an RNN model (90–95% detection accuracy; 88–92% Rasa classification accuracy). I also built an Azure-hosted continuous learning pipeline and a React + Flask + MySQL stack to support hybrid CI/CD. For speech-focused NLP, VocalEase converts stammered voice to fluent output using OpenAI Whisper with Flask-SocketIO streaming and gTTS synthesis.

I further demonstrated end-to-end product thinking with PersonalFitX (Top 8, Smart India Hackathon 2024), using Gemini AI with dynamic prompt engineering and a Django backend for real-time progress dashboards. I’m also motivated by collaboration—I mentored and coached a student team to “Best Emerging Team” at a college-level hackathon.