Rhea Generative Framework services and integrations
Rhea Generative Client
Internal business operations
Powerful Features for Rhea Generative Client
By integrating the Rhea Generative client, businesses can unlock enhanced capabilities, including fine-tuned models and persona additions. These features allow businesses to tailor AI responses to match their specific requirements and further control AI-generated responses.
The Rhea Generative Client offers a range of powerful features that enhance AI offerings and interactions. These features include:
- Chat-like interactions: Engage in natural, conversation-like interactions with AI systems, making it intuitive and user-friendly.
- Code formatting: Easily include code blocks in messages without syntax highlighting, enabling seamless collaboration and sharing of code snippets.
- Code copy to clipboard: Effortlessly copy code snippets generated by AI to the clipboard, streamlining workflow.
- Persona creation and activation: Create and manage personas to define constraints for AI output, customising and fine-tuning generated responses according to specific needs.
- Persona library: Access a rich library of predefined personas, making it convenient to select and apply constraints aligned with desired output.
- Authentication configuration: Easily configure authentication URLs and tokens for secure and authorised access to AI services.
- Theme support: Personalise workspace with a variety of themes to suit preferences and create a visually appealing environment.
- Content editor with completions: Enjoy a comprehensive content editor with auto-completion capabilities, enhancing productivity and enabling effortless creation of AI-driven content.
- Live Chat Functionality for Productivity Enhancement
The Rhea Generative Client can be utilised by employees to enhance productivity. By serving responses trained on their own datasets or utilising the Persona capabilities and model fine-tuning, businesses can generate responses adapted to their requirements. This functionality is particularly useful for front-of-office/front-of-house services, such as customer-facing live chat.
Rhea Embeddable
External business operations
Live Chat services with the Rhea Embeddable Client: Empowering Businesses with AI Capabilities
The Rhea Embeddable Client offers live chat functionality, enabling businesses to engage with customers in real-time. Powered by AI, businesses can provide instant support and personalised interactions, enhancing customer engagement, improving satisfaction levels, and building strong relationships through AI-driven conversations.
Customisable Chat Interface for Natural Interactions
With its user-friendly messaging-like interface, the Rhea Embeddable Client enables businesses to engage in natural, conversation-like interactions with AI systems. Whether your chosen backend is OpenAI, Azure OpenAI, or HuggingFace LLMs, businesses can unlock the full potential of AI capabilities and receive intuitive and user-friendly responses.
Train, constrain and gather
Rhea supports the ability to ground her responses to only information you provide no deviations.
- Train Rhea to capture new leads and politely reject everything else.
- Constrain her knowledge to particular subjects or fields.
- Gather information for a range of use cases such as lead capture.
Rhea features and capabilities
Persona’s Module
Personas can be used with the AI integration to shape and contain responses to specific subject areas. This module allows you to use several predefined persona’s or to create your own persona’s.
Persona’s can be installed on a user or global context:
- User context means that the persona will be applied to that user’s response only.
- Global context means that the persona will be applied to all users’ responses across the entire system.
Persona’s enables Rhea’s responses to be restricted to certain topics; limit responses and ensure certain personality traits. She can be trained to sign post users to particular types of content. If you have the right access she can query private data
Includes a Chromium Extension that can integrate with the Rhea RestAPI to allow you to save, edit and retrieve persona’s directly from the browser.
The Chrome extension enables secure communication to Rhea without needing to be logged in to Luna.
Rhea’s Rest API
The Rhea Generative Framework offers a comprehensive Rest API that includes the following capabilities:
- Completions: Generate high-quality content based on user input and the AI model’s training data.
- Chat: Facilitate natural, conversation-like interactions with AI systems for a seamless user experience. Converse and interact with Rhea over many topics or utilise personas to create focused “assistants” in particular subject areas.
- Document Generation and enhancement: dedicated endpoints for document generation and enhancements. By supply document template data and client questionnaire data in specific formats Rhea can extrapolate and populate templates for review.
- Persona Controls: Create and manage personas to customise and fine-tune AI-generated responses, aligning with specific requirements.
- Instructive: Harass persona’s to generate content based on persona constraints, user input and the AI model’s training data.
Rhea Generative Framework enables rich and consistent framework for building with AI services such as OpenAI, Azure OpenAI, HuggingFace, self-hosted LLMs and more. Enabling businesses to harass AI LLM (large language models) for their needs.
Rhea provides tools and functionality to shape and constrain generated output to specific business functionality or needs.
Businesses desiring or requiring more private options can use the Rhea Generative Framework to host their own models with compatibility for Rest API following OpenAI’s API schema suitable as a drop in replacement.
Disclaimer & Notices
Please note that our API explorer and PWA client are both publicly accessible but consumption of the services requires authentication; access requires an invitation or interested parties can request access at present we are not providing a public signup method.
