Ascend AI

AI-powered interview coach helping students break into tech. Practice with realistic AI interviews, get detailed feedback, and land your dream job with confidence.

Comment

Inspiration The inspiration for Ascend AI came from witnessing countless talented students struggle with tech interviews despite having strong technical skills. Many students, especially those from non-traditional backgrounds or without extensive networking opportunities, face significant barriers when interviewing for tech positions. Traditional interview preparation methods are often expensive, inaccessible, or lack the personalized feedback needed for improvement. We realized that AI technology could democratize access to high-quality interview coaching, providing students with unlimited practice opportunities and detailed feedback that would otherwise require expensive one-on-one coaching sessions.

What it does Ascend AI provides a comprehensive AI-powered interview preparation platform that includes:

Personalized AI Interviewer: Meet Sarah, an AI coach who conducts realistic mock interviews tailored to specific job titles and companies Real-time Recording & Analysis: Advanced recording system that captures both video and audio with proper MIME type handling for seamless playback Comprehensive Feedback: Detailed analysis of answer structure, speech delivery, body language, eye contact, and communication clarity Transcript Generation: Real-time conversation transcription with webhook integration for accurate content capture Performance Analytics: Track improvement over time with detailed scoring and recommendations Flexible Customization: Custom interview instructions and evaluation criteria for specific roles Secure Storage: Integration with Supabase for secure recording and transcript storage User-Friendly Interface: Intuitive dashboard with session management and progress tracking How we built it Frontend: Built with React, TypeScript, and Vite for a modern, responsive user experience. Implemented with Tailwind CSS for consistent styling and Lucide React for icons.

Backend: Developed using Node.js with Express.js and TypeScript for type safety. Integrated multiple APIs including:

Tavus API: For AI-powered video conversations with realistic avatars Google Gemini AI: For intelligent feedback generation and persona customization Supabase: For database management, authentication, and file storage Key Technical Features:

Real-time webhook integration for conversation transcription Advanced MIME type handling for video/audio recording Client-side recording system with system audio capture Comprehensive error handling and retry mechanisms Row-level security (RLS) for data protection RESTful API design with proper validation Architecture: Monorepo structure with separate frontend and backend services, enabling scalable development and deployment.

Challenges we ran into Tavus API Integration: Implementing proper webhook handling for real-time transcript capture required extensive debugging and multiple API format adjustments.

MIME Type Issues: Resolving video recording upload problems where files were incorrectly identified as text/plain instead of video/webm, requiring careful blob handling and validation.

Real-time Transcription: Ensuring accurate conversation capture through webhook events while handling network timeouts and API failures gracefully.

Audio Recording Complexity: Implementing client-side recording that captures both system audio (AI voice) and microphone input simultaneously.

Authentication & Security: Implementing secure user authentication with proper email validation and row-level security policies.

Cross-platform Compatibility: Ensuring the recording system works across different browsers and operating systems with proper fallback mechanisms.

Accomplishments that we're proud of Seamless AI Integration: Successfully integrated multiple AI services (Tavus, Gemini) to create a cohesive interview experience Real-time Processing: Implemented robust webhook transcription system that captures actual conversation content Technical Problem Solving: Resolved complex MIME type and recording issues with proper validation and error handling User Experience: Created an intuitive interface that makes AI interview coaching accessible to students Scalable Architecture: Built a maintainable codebase with proper separation of concerns and type safety Security Implementation: Established secure authentication and data storage with industry best practices Comprehensive Feedback System: Developed detailed AI analysis that provides actionable insights for improvement What we learned AI API Integration: Gained deep understanding of working with conversational AI APIs and handling real-time data streams WebRTC & Recording: Learned advanced browser recording capabilities and audio/video stream management Webhook Architecture: Mastered real-time webhook processing and event-driven architecture patterns File Handling: Developed expertise in proper MIME type handling and file upload validation User Authentication: Implemented secure authentication systems with proper validation and error handling Database Design: Created efficient database schemas with proper security policies and relationships Error Handling: Built robust error handling and retry mechanisms for external API dependencies What's next for Ascend AI Short-term Goals:

Mobile App Development: Create native iOS and Android apps for on-the-go interview practice Advanced Analytics: Implement machine learning models for more sophisticated performance tracking Industry Expansion: Add specialized interview tracks for different tech domains (frontend, backend, data science, etc.) Medium-term Vision:

Company Partnerships: Partner with tech companies to provide role-specific interview preparation Mentor Integration: Connect students with human mentors for additional guidance and support Group Practice Sessions: Enable peer-to-peer mock interviews and group preparation sessions Long-term Impact:

Global Accessibility: Expand to support multiple languages and international job markets Career Pathway Integration: Develop comprehensive career guidance beyond just interview preparation Educational Partnerships: Collaborate with universities and coding bootcamps to integrate into curricula

Built With

  • apinpm
  • authtavus-api
  • css
  • cssode.js
  • eslint
  • express.jssupabase
  • for
  • git
  • google-gemini-ai
  • html
  • javascript
  • mime
  • postgresql)supabase-storage
  • prettierjwt
  • react
  • rls)supabase
  • row-level
  • security
  • storage
  • supabase
  • tailwind
  • type
  • typescript
  • validationwebhooks
  • vite
  • webrtc
  • with
Share this project:

Updates