HackStack

Project Info

Winner

ElevenLabs: Best Use of ElevenLabs (11M Credits + AirPods Max per team member [1st] & 8.11M Credits + Merch [2nd] & 5.11M Credits + Merch [3rd])

The Duck You Mean?

Devpost

Inspiration

We were motivated by the power of rubber duck debugging and the wisdom of Richard Feynman’s quote: "If you can't explain it simply, you don't understand it well enough." Too often, AI is used to give answers, but true learning happens when we explain concepts ourselves. We wanted to flip AI on its head. Rather than asking the duck for answers, you teach it, and it pushes back with evaluations, scores, and follow-up questions to deepen your understanding.

What it does

The Duck You Mean is an interactive web app that lets users pick a topic and engage in a back-and-forth conversation where AI plays the role of a curious student. Users must explain concepts, and the AI: Scores their explanations based on clarity, correctness, and depth Asks intelligent follow-up questions to test deeper understanding Provides feedback to refine and strengthen their knowledge We also created zone called "The Pond", where users can create Zoom Meetings to study with friends or chill out after completing a lesson. Featured APIs -ElevenLabs - facilitates the full conversation process from user mic input to duck voice generation -Zoom API - creates meetings for peer-to-peer study sessions -OpenAI API - generates the starting question based on the prompt, -Mistrel AI - supports topic ideation and question formation -Luma AI - powers duck thoughts through created images

Challenges we ran into

Fine-tuning AI’s response evaluation—balancing depth of feedback while keeping interactions engaging Generating meaningful follow-up questions without redundancy Integrating voice interactions smoothly for an immersive experience

Accomplishments we're proud of

Successfully flipped AI’s role from answering questions to being the student Developed an engaging voice-driven AI tutor experience Created an effective learning tool that enforces deeper understanding Seamlessly integrated multiple AI-powered services

What we learned

Teaching a concept cements learning better than passive studying Thoughtful follow-up questions are key to uncovering misunderstandings Conversational AI can enhance education beyond just answering queries

What's next

Learning History to intelligently track user progress over time Gamification: Achievements, streaks, and leaderboards Mobile app version to promote learning on the go

Analysis

Compare with all teams

View

Metric

Figures cover GitHub contributors during the hackathon window. A co-authored commit counts in full for each author, so per-member totals add up to more than the whole-team figures.

Technology

Found in codeClaimed only
  • CSSIn code
  • FirebaseIn code
  • JavaScriptIn code
  • Next.jsIn code
  • OpenAIIn code
  • ReactIn code
  • Tailwind CSSIn code
  • TypeScriptIn code
  • Mistral AIClaimed

8 of 9 appear in the indexed code. 1 claimed on Devpost could not be matched to code, which may simply mean the tool leaves no trace in the repository.

AI coding agents

No AI coding agent signals were found in this repository.

Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.

Codebase size

Source size

44 KB

Source files

21

Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.

0 stars