Hi, I'm Aedwin Bobby — a computer engineering student building AI projects from Kerala

Greetings, I am Aedwin Bobby, a second-year Computer Engineering student at Mangalam College of Engineering in Kottayam, Kerala, India. My passion lies in crafting practical AI and machine-learning projects that solve real-world issues. This brief introduction aims to shed light on the projects I have been working on.

Most of my projects begin with identifying a real-world problem, which eventually transforms into functional software. One such project is the AI-Powered Network Intrusion Detection System, utilizing machine learning to identify malicious network traffic. Another example is AirSense-AI, an AI-AQI-Predictor that employs scikit-learn models to forecast pollutant gas levels (CO, NO₂, SO₂, O₃, PM2.5, PM10) and air-quality categories from AQI, complete with a user-friendly Gradio web interface.

Another innovative project of mine is the Hybrid AI Chiller Controller, which integrates reinforcement learning (PyTorch) with physics-based HVAC rules to suggest the most suitable chilled water supply temperature for large buildings, thereby minimizing energy waste. The Web Vulnerability Scanner is an AI-powered scanner and DevSecOps platform, featuring 15 modules aimed at bolstering cybersecurity.

One project close to home is the Kerala Flood Risk Predictor. It is a gradient-boosting model for Kottayam trained on 3,195 days of authentic rainfall (CHIRPS) and river-discharge (GloFAS) data. This project resonates with me, given the recurrent flood risk faced by Kerala during monsoon seasons.

Throughout my journey, I am continuously learning Python, machine learning, and the process of deploying AI that serves a practical purpose. Additionally, I am learning the less glamorous aspects that lend credibility to projects, such as creating clean READMEs, conducting honest evaluations, and understanding the constraints of models.

This experience has been invaluable, as I once witnessed a flood model's initial failure during real-world testing, and delving into the reasons behind it proved to be far more enlightening than the initial build.

For more information about my work, you can visit my website at https://aedwinbobby.github.io/, explore my GitHub profile at https://github.com/aedwinbobby, or connect with me on LinkedIn at https://www.linkedin.com/in/aedwinbobby. I am enthusiastic about internships and collaborations in the AI/ML domain and would be delighted to hear about any interesting projects you are working on.