Skaill

• 

  Scale your skills

Inspiration

Skaill was inspired by the growing need for efficient and personalized skill assessment in a remote-first world. We aimed to create an engaging and objective way for individuals to evaluate their expertise and receive actionable feedback, moving beyond traditional static tests to a dynamic, conversational experience.

What it does

Skaill provides an AI-powered video interview platform for skill assessment. Users can sign up, specify the skills they wish to be assessed on, and then engage in a live video conversation with a dynamically generated AI interviewer. The AI asks tailored questions, listens to responses, and provides real-time feedback. After the interview, users receive a comprehensive assessment report, including an overall score, detailed feedback for each skill, identified strengths, and areas for improvement.

How we built it

The application is built using React for the frontend, providing a responsive and interactive user interface. Daily.co powers the real-time video and audio communication, enabling the live interview experience. The core AI interviewing capabilities are driven by the Tavus API, which allows for dynamic persona generation and conversational AI. User authentication and data persistence (for user profiles, skills, and assessment results) are handled by Supabase. Styling is managed with Tailwind CSS and Shadcn UI components, ensuring a modern and clean design.

Challenges we ran into

One of the primary challenges was seamlessly integrating multiple real-time APIs, specifically Daily.co for video conferencing and Tavus for AI conversational capabilities, ensuring smooth data flow and synchronization between them. Managing the complex state of the interview flow, including countdown timers, question progression, and AI responses, also presented a significant challenge. Additionally, handling media device permissions and ensuring cross-browser compatibility for video and audio streams required careful attention. We also encountered and resolved several TypeScript errors related to type safety and unused variables during development, which improved code robustness.

Accomplishments that we're proud of

We are proud of successfully creating a fully functional, real-time AI video interview experience. The ability to dynamically generate AI personas based on user-selected skills is a key accomplishment, offering a highly personalized assessment. We're also proud of the intuitive user interface, the robust integration of video and AI, and the comprehensive, actionable feedback provided to users, which truly helps them understand and improve their skills.

What we learned

Through this project, we gained valuable insights into building real-time communication applications and integrating advanced AI services. We deepened our understanding of managing complex application states in React, handling asynchronous operations, and optimizing performance for video-intensive applications. We also learned the importance of robust error handling and user feedback mechanisms when dealing with external APIs and real-time interactions.

What's next for Skaill

For Skaill, the next steps include:

• Implementing more advanced AI feedback mechanisms, potentially including sentiment analysis and non-verbal cue assessment.
• Expanding the library of skills and question types to cover a broader range of professional domains.
• Introducing a progress tracking feature for users to monitor their skill development over time.
• Exploring integrations with learning platforms to recommend tailored courses or resources based on assessment results.
• Enhancing the AI persona customization options to allow users to select different interviewer styles or personalities.

Built With

• javascript
• raven-0
• react
• supabase
• tavus