About the Hackathon: EchoSphere: Agora Conversational AI Hackathon Build AI That Speaks, Listens, and Acts. EchoSphere is a hybrid AI hackathon where developers, students, startups, and AI enthusiasts come together to build the next generation of real-time conversational AI applications using Agora Conversational AI. This isn’t another chatbot hackathon. Your challenge is to create voice-native AI experiences that can hold natural conversations, handle interruptions, remember context, call external tools, and take meaningful actions in real time. Participants can choose from multiple real-world problem statements, including AI sales agents, classroom co-teachers, incident commanders, multilingual customer support, and adaptive AI interview panels—or build an innovative solution within the Open Innovation Track that showcases the power of real-time Voice AI. Throughout the hackathon, you’ll have access to workshops, technical mentorship, office hours, and guidance from industry experts. The event culminates in an offline Grand Finale in Delhi, where shortlisted teams will demonstrate their working prototypes live before an expert judging panel. What You’ll Build: Real-time Voice AI applications Conversational AI agents AI agents with tool calling Multilingual voice experiences Human-in-the-loop AI systems Production-ready AI prototypes Tech Stack: Use Agora Conversational AI as the core real-time voice layer, along with any LLMs, speech models, vector databases, cloud platforms, APIs, automation tools, or backend technologies of your choice. Why Participate? Build real-world Voice AI applications Learn from AI and industry experts Receive mentorship throughout the hackathon Network with developers and founders Showcase your project in the offline Grand Finale Win exciting prizes and recognition Whether you’re an experienced AI engineer or just starting your journey with conversational AI, EchoSphere is your opportunity to build innovative, production-oriented Voice AI experiences that solve real-world problems. Participation Guidelines: Team Size: Participants may participate individually or in teams of 2–4 members. Eligibility: Open to students, working professionals, developers, founders, and AI enthusiasts. One Team per Participant: A participant may be part of only one team. Team Changes: Changes to team members will not be permitted after the submission deadline without approval from the organizing team. Project Requirements: Projects must use Agora Conversational AI as a core component of the real-time voice experience. Teams may use any LLM, speech model, vector database, cloud service, API, framework, or backend technology alongside Agora. Projects should address one of the official problem statements or participate under the Open Innovation Track. The submitted solution must include a working prototype or functional demo. Projects should demonstrate meaningful conversational AI capabilities rather than functioning only as traditional text-based chatbots. Teams may use open-source libraries, APIs, pretrained models, and publicly available datasets, provided they comply with applicable licenses. Existing projects may be used as a foundation, but significant development and innovation must take place during the hackathon and should be clearly disclosed. Submission Requirements: Each team must submit: Project name and short description Selected problem statement/track Team member details GitHub/source-code repository Working demo or deployment link, where applicable Demo video Architecture/technology stack Explanation of how Agora Conversational AI is integrated Any additional information requested by the organizers Incomplete or inaccessible submissions may not be considered for evaluation. Grand Finale: Shortlisted teams will be invited to the offline Grand Finale in Delhi. Finalists must demonstrate a working prototype live before the judging panel. At least one team member must be available to present the project in person unless the organizers approve an exception. Teams may be required to participate in a short presentation, live demo, and Q&A session. Evaluation Criteria: Projects may be evaluated based on: Innovation & Creativity Use of Agora Conversational AI Technical Implementation Quality of Voice/Conversational Experience Real-world Impact & Usefulness Product Readiness & Scalability Live Demo & Presentation The judges' decisions will be final and binding. Code of Conduct & Fair Play: Participants must maintain professional and respectful behavior throughout the hackathon. Plagiarism, unauthorized copying of another team's work, misrepresentation of project ownership, manipulation of submissions, or other unfair practices may result in immediate disqualification. Teams are responsible for ensuring they have the right to use all code, datasets, APIs, models, media, and other third-party resources included in their projects. Intellectual Property: Teams retain ownership of the original intell