Introducing Prism - VideoSDK Agents V1.0.0
Blog post from Video SDK
Prism: Agents V1.0.0 represents a comprehensive overhaul of the VideoSDK AI voice pipeline, addressing limitations found in previous versions. The update introduces a unified Pipeline class that replaces the previous CascadingPipeline and RealtimePipeline, allowing for greater flexibility and interoperability between components such as speech-to-text (STT), language models (LLM), and text-to-speech (TTS) systems. This new architecture supports three modes—Cascade, Realtime, and Hybrid—enabling customization and integration with various models and tools, while also enhancing observability with built-in metrics, logging, and tracing. The system is designed to run concurrently, reducing latency and improving responsiveness, and includes a robust hooks system for data interception and transformation at different stages. The release includes 22 production-ready templates for common use cases and maintains compatibility with existing tools and lifecycle management features, providing a scalable and flexible framework for AI voice agents.
| Trend | Post Mentions | Total Month Mentions | Posts | Companies | MoM |
|---|---|---|---|---|---|
| LLM | 26 | 6,889 | 1,263 | 265 | -9% |
| Real-time | 19 | 7,450 | 1,704 | 292 | -47% |
| MCP | 7 | 7,956 | 795 | 196 | +24% |
| Voice AI | 4 | 3,611 | 281 | 50 | -5% |
| Observability | 3 | 4,900 | 921 | 200 | +5% |
| AI Agents | 1 | 5,835 | 1,407 | 272 | -21% |
| OpenTelemetry | 1 | 1,168 | 142 | 46 | +24% |
Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.