Developed and delivered a real-time audio transcription feature for the open-edge-platform/edge-ai-libraries repository, addressing a key need for live analytics and captioning workflows. The work involved integrating Whisper-based transcription from OpenVINO GenAI into the DLStreamer pipeline, enabling transcription from multiple audio sources in real time. Collaborated closely with another contributor to ensure robust implementation and cross-team alignment. Focused on validating both streaming latency and transcription accuracy, the solution supports faster time-to-insight and improved accessibility for end users. The project leveraged expertise in C++ development, audio processing, and GStreamer, with all changes completed and merged within a focused development cycle.