Inspiration
Our inspiration for Talk It Out came from personal experiences with conflict resolution scenarios, where we often felt unprepared to handle real-life situations effectively. We recognized a significant gap between theoretical knowledge and practical application opportunities for developing these essential soft skills.
What it does
Our web app simulates realistic conflict resolution scenarios where users interact with an AI co-worker through voice. As you speak into your microphone, our program transcribes your message and feeds it to our carefully prompt-engineered AI, which responds dynamically to create an authentic conversation experience. Moreover, our AI also considers nonverbal gestures too, as communication isn’t solely about words, via facial recognition software through the webcam. After each session, the AI provides specific feedback based on established frameworks like Nonviolent Communication and the Thomas-Kilmann Conflict Management Model.
How we built it
We built Talk It Out using Next.js and React for our frontend development, while integrating OpenAI and ElevenLabs APIs to create realistic simulations that help users improve their social and conflict resolution skills in a practical setting.
Challenges we ran into
Our main challenge was fine-tuning prompts to ensure the AI would engage in natural conversation while providing actionable, constructive feedback that users could immediately apply to improve their conflict resolution abilities.
Accomplishments that we're proud of
We're proud of rapidly bringing this concept to life and creating a tool that addresses a real need for practical soft skills development in a safe, accessible environment.
What we learned
Through this project, we gained valuable insights into prompt engineering complexities and learned how to integrate multiple APIs seamlessly to create a cohesive user experience that feels natural and educational.
What's next for Talk It Out
We plan to expand our scenario library to cover diverse conflict types and personalities, and implement a backend system that enables personalized learning paths by tracking user progress and tailoring feedback over time.
Log in or sign up for Devpost to join the conversation.