Brajakishore Behera@brajakishorebehera007
1
Events attended
1
Submissions made
India
1 year of experience
About me
Hi there! I am Brajakishore Behera, a passionate technologist and developer based in India. I am deeply fascinated by the rapid evolution of Artificial Intelligence and its potential to solve complex, real-world problems. My journey in tech is driven by a strong curiosity to understand how things work under the hood and a relentless desire to build innovative solutions from the ground up. As an active participant in the tech community, I thrive in dynamic environments like hackathons where creativity meets rapid prototyping. I am constantly exploring the latest advancements in AI, machine learning, and modern software development, always looking for new tools and frameworks to add to my skill set. Whether it is experimenting with new language models, building generative AI applications, or optimizing backend logic, I truly enjoy the entire lifecycle of product creation. Beyond writing code, I am a huge advocate for collaboration and continuous learning. I believe that the most groundbreaking ideas emerge when diverse minds and perspectives come together to tackle a challenge. I am here to connect with like-minded innovators, form strong teams, and push the boundaries of what we can build over a weekend. When I am not coding, you can find me engaging with the community, sharing insights on tech trends, or brainstorming new project ideas. I am always open to networking, discussing potential collaborations, or just having a great conversation about the future of tech. Feel free to connect with me through my linked profiles—let's build something amazing together!
🤓 Latest Submissions
SafeGuard AI
Project Description: SafeGuard AI Overview SafeGuard AI is an advanced multi-agent framework designed to revolutionize industrial infrastructure monitoring. Developed for the AMD Developer Hackathon, the system addresses the critical need for real-time, autonomous anomaly detection and rapid emergency response in large-scale facility management. The Problem Traditional infrastructure monitoring often relies on manual oversight or simple threshold-based alerts, which can lead to delayed responses or "alarm fatigue." In high-stakes environments, every second counts when preventing hardware failure or safety hazards. The Solution SafeGuard AI utilizes the CrewAI framework to orchestrate a specialized team of autonomous agents that work in tandem to monitor, analyze, and react: Sensor Analysis Agent: This agent continuously processes streams of simulated IoT data—tracking variables like temperature, vibration, and pressure. It identifies subtle patterns that indicate potential risks before they escalate. Safety Protocol Agent: Upon detection of an anomaly, this agent instantly takes over to draft actionable emergency repair protocols. This ensures that human responders have immediate, expert-level guidance the moment an issue arises. AMD Integration & Technical Stack The system is built with high-performance scalability in mind. The architecture is conceptually optimized to leverage the massive parallel processing capabilities of AMD Instinct MI300X GPUs. By utilizing the ROCm (Radeon Open Compute) software stack, SafeGuard AI is designed to scale across thousands of concurrent sensor streams with minimal latency, providing a robust solution for enterprise-level industrial applications. Key Features: Autonomous Orchestration: Multi-agent coordination via CrewAI. Proactive Safety: Transitions from reactive alerts to proactive repair drafting. Scalable Architecture: Optimized for the ROCm compute stack and AMD hardware.
10 May 2026
