Owlyn

Description

A real-time multimodal AI platform for technical interviews and developer assistance that can see, hear, and reason about the user’s live workspace using specialized AI agents for voice interaction, visual analysis, code reasoning, and workspace monitoring, with LiveKit WebRTC supporting real-time audio and video interaction.

Highlights

• Co-developed a real-time multimodal AI interview platform that analyzes spoken responses, video signals, and developer workspace activity to assess candidates during technical sessions.
• Designed an Electron client and backend services that coordinate specialized AI agents for voice interaction, visual analysis, and code reasoning across technical interview sessions
• Built real-time audio and video pipelines with LiveKit WebRTC, enabling live interaction between candidates and the AI interview system.
• Implemented real-time interview proctoring with vision-based monitoring and live human oversight, allowing proctors to observe candidate activity and intervene during technical sessions.