Abra AI: AI Mock Interviews

The new era of recruiting. Instant feedback with detailed analytics.

(live demo)

Table of Contents​

📝 About
💻 How to Build
🚀 My Next Steps
🔧 Tools Used
👤 Contact

📝About​

Abra AI utilizes ML speech-to-text models to deliver mock interviews wth instant feedback and criticism of your performance. Choose from a variety of questions commonly used during technical and behvaioral interviews (FAANG, Spotify, Uber, McKinsey, and more). Ace your interview using data-driven insights.

Receive unbiased feedback and support from our community of experts and peers.

Input your own interview questions and create dynamic interview scenarios to practice for your big day.

Get instant feedback analytics from our AI and hone your interviewing skills.

💻How to Build​

• DeepSpeech (speech recognition) + spaCy (NLP) for analyzing verbal logic.
• OpenCV computer vision for facial analysis to detect emotional cues.
• FFMPEG was used to process the audio recordings. WebSockets used for low latency video streaming between interviewer and interviewee.
• OpenAI GPT-4 to process text responses and feed data-driven user data to FFMPEG, then send response back in avatar persona style to mimic interview process.
• AWS Lambda infrastructure used for scalable deployment.
• GraphQL for API. Helpful for hosting company data and question sets.
• Frontend website built on NextJs.

🚀My Next Steps​

Expecting to launch this soon and add user authentication and waitlist forms. Stay tuned 👀.

🔧Tools Used​

ML

APIs / Web tools

Framework/Deployment

👤Contact​