Panel 2 - AI Applications - LogFlowAI - Transforming DevOps with Automation, Real-Time Analytics & AI

LogFlowAI is an AI-powered DevOps tool that uses predictive analytics and real-time monitoring to proactively manage system performance. Our research lies at the intersection of agentic-AI, DevOps, and machine learning, specifically exploring how neural networks and statistical time series models can forecast system failures from log data. We’ve employed streaming data processing via Kafka, Zookeeper, and GPU-accelerated computations, and the project is ongoing, currently focused on extending our methods and refining the predictive accuracy of our models under guidance of Professor Riondato as well. This project won HackMIT 2023 (Y Combinator Challenge - 3rd Place), and was recognized by judges from NVIDIA, Meta, OpenAI, Walmart, Liquid AI, Tesla, and more. I also received a special Prize from Tesla - 48-hour demo drive for groundbreaking AI DevOps innovation.