Ai-Powered Podcast Transcription and Semantic Annotation System Development

We are seeking an experienced AI developer or team to build a comprehensive system for automated, high-quality transcription and semantic annotation of audio podcasts. The primary goal is to transform raw audio into structured, searchable data.

The system will involve the following key components and functionalities:

1.  Audio Transcription: Implement and integrate Whisper for accurate transcription of audio podcast files.
2.  Semantic Annotation: Develop and fine-tune a custom Large Language Model (LLM), such as LLaMA or Mistral, to process the transcribed text.
3.  Content Summarization: The LLM will be used to generate concise summaries of the podcast content.
4.  Key Topic Extraction: Identify and extract the main themes and key topics discussed within each podcast.
5.  Mood and Theme Tagging: Automatically tag the audio content with relevant moods and themes to enrich metadata.
6.  Searchable Database Integration: Store all transcribed text, summaries, extracted topics, and tags in a searchable database for easy retrieval and analysis.

Expertise in Hugging Face transformers and LoRA (Low-Rank Adaptation) for efficient LLM fine-tuning is essential. The ideal candidate will have a strong background in natural language processing, machine learning, and audio processing, with proven experience in building robust AI-driven solutions.