About the Pryon Generative Engine
The Pryon Generative Engine offers developers a flexible, reliable, and secure toolkit for implementing customizable generative solutions, offering a framework for rapid implementation of retrieval augmented generation solutions through a simple point and click LLM integration.
The Generative Engine features the ability to customize prompts and configure LLM orchestration to solve customer problems, fine-tuning LLMs leveraging the Pryon Ingestion Engine to facilitate workflow orchestration. It can be installed at your location, or in a private cloud environment in order to mitigate the risk of sending IP to public LLMs.
Generative API categories
The major categories of Generative APIs are: Exchange and Feedback.
Exchange Events
URI: /api/conversation/v1/exchange-events/sse
ExchangeEvent service provides methods for creating and managing exchange events. The implementation uses a Large Language Model (LLM) to create task/chat-oriented models. The tasks include interpreting question query, chit-chat detection, and question/answering task based on given contexts from domain specific models. These models are used to create a complete end-to-end Question/Answering system. The service uses server-sent events (SSE) to send different events during the process.
The Exchange API is used to initiate or continue a conversation with the conversation exchange service. Each call performs one synchronous round-trip by sending a text or audio request to the service and receiving the text and optional audio response.
Generative Feedback
URI: /api/conversation/v1/generative-ratings
The Feedback API is used to provide, and manage, user feedback for a given exchange and response answer. This allows for user input as to the accuracy of the question/answer pair and is used as a basis for the Verified Answer feature of the Pryon Service.