Proctor mode ensures the integrity of take-home assessments by monitoring the session for any suspicious activities via signals from the webcam, screen capture, and other sources, and presenting the findings in a report. It's intentionally designed not to impact the candidate experience unless there's suspicious activity.
Screen to Interview Identity Match compares candidate images from take-home assessments in our Screen product with the images from the video feed in our Interview product and flags if there's a mismatch.
The IDE now includes an AI assistant that mirrors real-world developer workflows. In take-home assessments, it works as a guarded assistant, providing syntax and debugging support while maintaining assessment integrity.
Advanced evaluation extends beyond evaluating the candidates on code correctness and optimality. It assesses the quality of the code written, how they worked with the AI, and other factors, and all of these are summarized and presented in the report.
In interviews, the IDE offers an unguarded AI environment that mirrors the real world. Candidates can use Agentic mode, get inline code assistance, and even choose their preferred AI model. This lets you see how they collaborate with AI in practice, the way they’ll code on day one.
Scorecard Assist uses AI to generate a structured summary from the interview session, analyzing a combination of the transcript and code playback. This is summarized in the rubric specified by the interviewer, helping them complete the writing of the scorecard more efficiently.