Rather than a separate "multimodal" analysis layer, Deepgram's Audio Intelligence features run alongside transcription on the same audio: summarization, topic detection, and sentiment analysis, so a single API call returns more than a flat transcript.
- Automatic summarization of calls or recordings
- Topic detection and sentiment analysis scored across the transcript
- PII and PCI redaction that can mask sensitive spoken data, such as card numbers, directly in the transcript
Use cases
Intro
- Synchronize calls with video contexts using Deepgram
- Correlate spoken terms with event logs via Deepgram
- Link audio transcripts to video frames in analytics
- Enhance customer journeys with multimodal signals from Deepgram
- Compliance and risk assessment using synchronized data from Deepgram
Closing