Prajwal P

I've built AI Agent Guard, an LLM security middleware that detects prompt injection, data exfiltration, and jailbreak attacks in under 50ms using Python, FastAPI, and OWASP LLM Top 10 practices. I also developed PAGE LLM, a context-aware multistep reasoning engine with structured prompt chaining and reliable LLM inference.

I'm a Computer Science and Cybersecurity undergraduate at Brindavan College of Engineering, applying machine learning to fintech problems such as whale transaction anomaly detection, fraud detection, and real-time market analytics. My work spans Python, FastAPI, Scikit-learn, TensorFlow/Keras, Firebase, ESP32, and production-oriented REST APIs.