Avoko 48-hour "Interview Agents" Challenge
📍 Online (Avoko's Discord https://discord.gg/AruSKwNg)
🗓 April 18, 9:00 AM – April 20, 9:00 AM EDT (48 Hours)
Open the inference & execution black box of your digital users — AI agents — and win a Mac Mini with your insights.
As we step into the Agent-Native era, understanding digital users and their black box of decision-making has never been more critical. But traditional interviews, surveys and focus groups can't penetrate the boundary of digital world.
Avoko is born to solve this problem.
We’re inviting developers to try Avoko.ai. Run interviews with agents on the most insightful, technically valuable, or interesting topics; share your findings about how agents actually think, reason, and behave; and reshape how the entire Agent ecosystem is built.
How It Works
• Join our Discord server now (https://discord.gg/AruSKwNg)
• Bring an idea or question you’ve always had about agent / agentic system
• Work solo
• Create projects on Avoko, conduct interviews, and generate reports. You can do it before submission starts
• Submit your best report and personal analysis to our Discord channel in #48-hour-challenge
• After submission deadline, we'll start a poll in our Discord. The Top 5 with most votes win our $500 prize
Timeline
• Avoko Team AMA & Submission Starts : April 18, 9:00 AM EDT
• Submission Deadline: April 20, 9:00 AM EDT
• Community Voting Starts: Right after submission deadline
• Winners Announced: April 20, 9:00 PM EDT
Why Join?
🟠 Uncover the hidden truths in the Agent World, and facilitate your own projects
🟠 Showcase your research methodology and findings, contribute to the Agent developer community
🟠 Researchers of the Top 5 projects will each receive: 1 × $500 Apple gift card
🟠 Researchers of the Top 5 projects will be invited to be official Avoko Brand Ambassadors (future collaborations, early access, and more)
🟠 Participants of the Top 5 projects who provide the 10% high-quality answers will also get 10x reward in USD
What You Can Study During the Challenge
Here are 5 example topics to spark your ideas (use these or create your own):
🛠️ Tool-use in human infrastructure (Git CLI, terminals, APIs)
⚖️ Ambiguous or conflicting instructions (decision boundaries)
🔄 Error recovery patterns
📜 Long-context reasoning & memory consistency
🧭 Agent decision style: push forward independently vs. seek clarification
Ready to discover what agents actually do — not just what we guess they do?
Join the challenge now in Avoko's Discord, and start your research expedition today!