Mockingbird

Mockingbird is an AI interview simulator that tailors mock technical screens to specific job postings. It provides real-time feedback to help developers ace the job.

Comment

The Inspiration

The idea for Mockingbird was born out of a common frustration in the job market: the "experience gap." Many candidates have the technical skills but freeze during the high-pressure environment of a live interview. We noticed that while tools like ChatGPT can generate interview questions, they lack the rhythm and pressure of a real conversation. We wanted to build something that felt less like a chatbot and more like a human sitting across the table, one that actually knows what a specific technical role is truly looking for.

How We Built It

Mockingbird is built using a diverse stack that combines a responsive frontend with a robust backend to handle the complexity of real-time voice interaction.

  • Frontend: The interface is built with HTML, CSS, and JavaScript, ensuring a lightweight and accessible user experience.
  • Backend: The core logic is powered by Python, which handles the heavy lifting of processing job data and managing the AI logic.
  • Voice Integration: To achieve the "speaking" aspect of the tool, I integrated APIs that allow for seamless Speech-to-Text and Text-to-Speech conversion, enabling the AI to converse naturally with the user.

Challenges Faced

  • Latency ("The Awkward Pause"): In early builds, the delay between a user finishing a sentence and the AI responding was a major hurdle. We had to optimize the communication between the Python backend and the JavaScript frontend to ensure the conversation felt fluid.
  • Job Posting Parsing: Job descriptions vary wildly in format. We had to refine the backend logic to reliably extract the most important skills and requirements regardless of how the posting was structured.
  • Cross-Language Integration: Coordinating the data flow between the JavaScript-driven UI and the Python-driven AI engine required careful state management to prevent "drifting" during the interview phases.
  • Unimplemented Features: We had more functionality in the backend for more in depth grading and scoring of the users answers and rewriting them. However, we we ran out of time and did not fully get the front end implementation added for these features.

What We Learned

Building Mockingbird was very different for all of us as we had not met each other before this hackathon. We learnt that communication is vital and being organized is the most important thing.

Built With

Share this project:

Updates