LLM Feedback Loops || Voice AI Hackathon
Overview
Join us for a hackathon focused on advancing Voice AI through intelligent feedback loops and evaluation systems. Using ElevenLabs’ voice synthesis and LLMs, we’ll build adaptive voice agents that can self-diagnose and improve their own system prompts over time.
Evaluating LLM systems is a major challenge—and we're tackling it head-on. Whether for interviews, data collection, or dynamic conversations, let’s create voice agents that don’t just talk, but continuously self-improve.
Who Should Join?
Engineers, prompt designers, LLM researchers, UX thinkers, conversation designers, and anyone curious about voice interfaces and AI feedback loops.
Come with a project idea, or find collaborators at the event. Solo? We'll help with team formation after kickoff.
Examples:
Expert Interviewer: A voice agent that questions domain experts to capture both quantitative metrics and rich qualitative insights, then refines its questioning strategy through feedback‑driven prompt tuning.
Learning Buddy: A voice companion that adapts tone, pacing, and complexity to the user’s comprehension level—self‑diagnosing and improving its system prompts from interaction history.
Sponsors
ElevenLabs – Voice AI reimagined.
Prizes
- 3 months free of the Creator tier for all attendees
- 1.1+ million characters for each member of the overall winning team (3 months of the Pro tier)
- 11+ million characters for the "Best project built with ElevenLabs" for each team member
Judging
Looking for systems:
- Adaptive Intelligence: Self-Evolution
- Conversational Craft: Human Connection
- Real-World Impact: Practical Innovation
Judges
Tobias Hanl (Rilla)
Sunjana Ramana
Max Lynch (11Labs)
Hari Prasad Renganathan
Sarthak Singh Chauhan & Pratham Bansal (Lanesurf)
Schedule
8:30am - Doors Open (coffee, mingle)
9:30am - Hackathon Starts (talks)
10:00am - Team Assignments, Coding Begins
1:00pm - Lunch
4:00pm - Submissions
4:15pm - Demo
5:00pm - Judging & Prizes
Contact us here.