Disha Jain@disha_belieber
1
Events attended
1
Submissions made
India
1 year of experience
About me
Computer Engineering student and aspiring AI Engineer passionate about building real-world products with Artificial Intelligence, Machine Learning, and Full-Stack Development. I enjoy solving complex problems through technology and have experience developing AI-powered applications, privacy tools, and data-driven solutions. I actively participate in hackathons, contribute to open-source projects, and continuously explore emerging technologies. My goal is to create impactful products that improve people's lives while growing as a software engineer and entrepreneur.
🤝 Top Collaborators
🤓 Latest Submissions
Cognexa – Hybrid AI Routing Platform
Cognexa – Intelligent Hybrid AI Routing Platform Cognexa is a hybrid AI routing platform developed by Team Horcrux for the AMD Developer Hackathon (ACT II). Instead of sending every request directly to a large language model, Cognexa first analyzes the incoming prompt to determine the most efficient execution path. The platform classifies user requests into categories such as mathematics, sentiment analysis, named entity recognition (NER), factual knowledge, summarization, and code generation. Lightweight tasks are executed locally using specialized handlers, while more complex or knowledge-intensive requests are automatically escalated to Fireworks AI. This hybrid routing strategy reduces API calls, lowers inference costs, improves response times, and optimizes token usage without sacrificing response quality. Cognexa also provides a transparent decision pipeline through an interactive dashboard. Every request is logged with its detected category, selected execution route, confidence score, latency, handler, and estimated token savings. A routing visualizer allows users to understand how the platform makes routing decisions, making the system easier to debug, monitor, and explain. The project is built with a modern full-stack architecture using React, Tailwind CSS, FastAPI, and Python. The frontend offers a responsive dashboard with real-time analytics, while the backend manages classification, routing, local execution, and external model integration. Fireworks AI is used only when local processing is insufficient, enabling efficient resource utilization. By combining intelligent routing, local execution, explainable AI workflows, and cost-aware inference, Cognexa demonstrates how hybrid AI systems can deliver faster responses, lower operational costs, and scalable deployment for enterprise AI assistants, customer support, developer tools, and educational applications.
13 Jul 2026
