Show HN: Open-source customer-facing agents (github.com)

Radioso has launched an open-source platform designed for customer-facing agents that facilitates guided interactions using large language models (LLMs). This platform allows developers to craft conversational experiences powered by their own documents, ensuring responses are grounded and cited accurately. Significantly, Radioso introduces "guided autonomy," where developers can set rules that dictate agent behavior without needing to map out every possible conversational path. This flexibility is crucial for enhancing customer support models while maintaining user control. Key technical features include the ability to run Radioso Cloud or on local infrastructure, avoiding vendor lock-in with multi-provider support for models like OpenAI, Anthropic, and Gemini. The platform operates on an API-first architecture, offering capabilities like routine management, directives for real-time adjustments (e.g., slowing down for anxious customers), and human handoff mechanisms for complex queries. Moreover, integration with tools such as Slack and email response frameworks positions Radioso as a versatile solution for businesses looking to enhance their customer interaction strategies through AI.