We won 1st Place at ElevenLabs Hackathon
We just won the ElevenLabs Online Conversational Agent Hackathon — here's the story
What is the ElevenLabs Conversational Agent Hackathon?
- 2-hour online sprint. Each team had exactly two hours to design and ship a working voice agent live on the ElevenLabs platform. Took place on Wednesday, July 2, 3pm - 5pm UTC.
- Global event celebrating 1 million agents built on ElevenLabs, with partners like Lovable, Exa, n8n, Notion and others.
- Top prize: US $20 000 in ElevenLabs credits for 1st place (that's us 🎉).
Our idea: a "Voice Nurse" built in under 2 hours
We focused on healthcare because fast, accurate answers save stress — and sometimes lives.
Key abilities
What it does | Why it matters |
---|---|
Listens to a symptom description in natural speech | Removes the need to fill in long forms when you're worried. |
Asks the right follow-up questions | Mimics the logic of a triage nurse to narrow down the issue. |
Gives clear next-step guidance (self-care vs doctor visit) | Users leave the call knowing exactly what to do. |
Books an appointment or connects to a human line on request | One call, zero hand-offs. |
Technical Implementation
We stitched everything together with ElevenLabs' Conversational AI for real-time speech, a lightweight symptom-checker API. Total build time: 1 hour 59 minutes plus a quick demo video.
Step | What happens | Tech used |
---|---|---|
1. Caller speaks symptoms | ElevenLabs Conversational AI transcribes and voices the dialogue in real time. | ElevenLabs API |
2. Smart follow-ups | A lightweight symptom-checker API asks clarifying questions. | External API |
3. Guidance & hand-off | The agent tells the caller whether self-care is fine or a visit is advised. | Agent logic |
4. Instant booking | If needed, the agent opens a pre-filled calendar slot and confirms by SMS. | Lovable backend + Twilio |
Try it yourself
Click the live demo and ask about your symptoms in plain English:
👉 Launch the Voice Nurse
(Works best on Chrome/Edge with a headset.)
Youtube demo presentation:
What the judges liked
- Practical vertical. "We were really impressed by the product: you built and the integration it had with the voice agent. We're really excited to see the future of voice agents in healthcare so we're so happy to see what you built." — Angelo, ElevenLabs
- Polished hand-off. The smooth transfer to a human line showed a clear path to production use.
- Complete experience in two hours. From greeting to booking, everything ran end-to-end and live.
What's next for the project
- HIPAA-grade data handling. We're mapping out the compliance work needed for US deployment.
- Multilingual voice models. Moldovan Romanian and Russian come first, then wider EU languages.
- Pilot with local clinics. We're lining up two private practices for a limited beta this autumn.
- Open-source snippets. The triage flow and hooks will be published on GitHub after code cleanup.
Why this matters for Aichat.md
Winning the ElevenLabs hackathon gives us:
- • $20 000 in voice credits — enough capacity to serve thousands of patient calls while we refine the product.
- • Visibility in the AI-voice community, attracting partners and talent.
- • Proof that our team ships fast under real-world constraints.
This momentum feeds directly into our larger mission: building AI agents that take busywork off human teams so they can focus on high-value conversations.
Got feedback or want to run a pilot?
Email us at info@aichat.md — we'd love to talk.
22 July 2025