PolyAI has announced that its Agentic Dialog Platform is now open to all enterprise builders, enabling teams to create and deploy voice AI agents without the need for extensive contracts. The platform features Raven, a proprietary model designed specifically for dialog, which excels at managing complex interactions and maintaining context during challenging calls. This move aligns with a broader industry trend of making specialized AI tools accessible through self-serve trials, a strategy aimed at enhancing productivity for contact centers and operational teams, as highlighted by recent commentary on the importance of voice-native AI agents in addressing emotionally charged conversations.

Raven: Raven is PolyAI’s proprietary dialog model, designed specifically to manage complex, high-pressure conversations by keeping track of evolving user intent and context. In the news, Raven is highlighted as the default engine inside the Agentic Dialog Platform, with agent behavior encoded in the model weights rather than long prompts, while still allowing users to swap in other large language models if needed.
PolyAI: PolyAI is an enterprise-focused conversational AI company that builds lifelike voice and dialog agents for complex customer service scenarios across many industries and languages. In this news, PolyAI is opening its Agentic Dialog Platform and associated tooling to all builders, shifting its previously high-touch, large-contract technology into a broadly accessible, self-serve model with a free trial period.
Poly Agent Builder: Poly Agent Builder is a no-code tool within PolyAI’s ecosystem that lets users describe their use case in natural language and automatically generates a working dialog agent, including knowledge bases and conversation flows. In this context, it is positioned as one of two primary build paths on the newly opened Agentic Dialog Platform, promising production-ready agents in minutes for non-technical teams.
Agent Development Kit: The Agent Development Kit (ADK) is PolyAI’s developer-focused toolkit for building and deploying dialog agents using standard software engineering workflows and tools. In the news, it is presented as the code-centric counterpart to Poly Agent Builder, enabling developers to integrate with IDEs, use coding assistants, manage versions with Git, and deploy agents directly from the terminal.
Agentic Dialog Platform: The Agentic Dialog Platform is PolyAI’s end-to-end environment for designing, running, and managing production-grade dialog agents across voice and chat channels. In this announcement, the platform is being made self-serve and widely accessible, allowing any team to create, test, and deploy agents on the same infrastructure used by large enterprise customers.

Go_To_Market: Opening previously enterprise-only AI platforms via self-serve trials and developer toolkits has become a common strategy among leading AI vendors to accelerate adoption, expand beyond large contracts, and empower operational teams to experiment directly.
Voice_AI_Trend: Recent industry commentary highlights voice-native AI agents as a key productivity lever for contact centers and operations teams, with growing emphasis on systems that can handle complex, emotionally charged calls rather than simple FAQs.
Model_Architecture: Specialized dialog models like Raven are increasingly being differentiated from general-purpose chat LLMs by their ability to maintain stable context, handle interruptions, and orchestrate multi-step actions during live conversations.