Components

FasterWhisper

Speech inference component for transcription workloads in Runtime.

Agentic Friendly

Component Category

Inference / speech-to-text

Component Description

FasterWhisper is an optimized implementation of Whisper for efficient speech transcription workloads.

Why It Is Used

In BullSequana AI Runtime, FasterWhisper supports audio transcription use cases with lower latency and lower resource consumption than a baseline Whisper deployment.

Learn More

Deployment notes

FasterWhisper is not deployed as a standalone component. It is registered as a model server engine in KubeAI, which can manage FasterWhisper pods dynamically based on Model custom resources.

Interacts With

  • KubeAI, as part of the speech-serving and model execution layer.
  • Model Installer and speech-oriented serving workflows, which are responsible for packaging and deploying transcription models.

vLLM

High-throughput LLM inference engine in the Runtime layer.

Kyverno

Kubernetes policy engine for validation, mutation, and resource generation.

On this page

Component CategoryComponent DescriptionWhy It Is UsedLearn MoreDeployment notesInteracts With