Best-in-class
Architecture
A highly performant RAG architecture that comes with the advanced features required for enterprise-grade deployments
Advanced Disambiguation
Handle complex queries with multi-intent resolution and disambiguation flows
Conversational Memory
Enhance dialogue dynamics and expand NLP understanding accuracy
Anti-hallucination Pipeline
Transform AI reliability with trustworthy business outputs
Knowledge Gaps Pinpoint
Automatic determination of missing intents of your knowledge sources
Knowledge Self-expansion
Automatic domain alignment by augmenting a dataset based on your documents
Prompt-hacking Protection
Secure your AI to safeguard your enterprise information
All-in-one software stack for generative chats
Chat Widget
Customizable off-the-self widget with history, rating, resizing management that easily integrates in your favourite framework so you can talk with your chatbot from your website. Alternatively use the frontend API to integrate with any external app
Back-end
It takes care of everything you don’t want to (components connections, session storage, datasets and LLM registration, as well conversation Finite-State-Machine)
Conversation flows SDK
It allows you to expand your LLM with custom conversational flows, by executing complex Finite State Machine behaviours against a remote custom logic over RPC protocol
AI Engine
It takes care of local and cloud AI-related tasks, including executing LLMs inferences, parsing contents, creating embeddings, generating utterances, indexing contents and training custom models
Comprehensive Management
RAG pipeline on any enterprise documents
Control all aspects of the RAG pipeline and tune your LLM on any enterprise documents:
- Ingest your Knowledge Items into the RAG dataset (via CSV/PDF sources or API)
- Select your LLM model of choice (Claude, OpenAI, Mistral, Llama,…)
- Define and train your RAG retriever model against an auto-generated RAG dataset
- Create complex conversations involving multiple RAGs
- Automatic handling of chitchat intents and conversations
End-to-end service-centric dashboard
Control all service capabilities through the admin back-office to manage the chat settings and service aspects:
- Dashboard to quickly glance over your active chatbot instances KPIs
- Simplified labelling to evaluate live questions and responses
- End-to-end AI settings for the entire RAG pipeline (Knowledge items, retriever, prompt, LLM)
- Personalization of the chat widget UX/UI aspects (features, authentication, localization, theming)
- Get the audit trail of all configuration changes via the Task change history
Build for
Best-of-breed software stack
$ 6.99 USD
Foundation based on Django framework and the JS widget pluggable in the component-based library of your choice
Flexible deployment options
$ 10.99 USD
Be in complete control over your data privacy and security by choosing your deployment models and architecture
Almost everything for free
$ 14.99 USD
This open-source project comes without license fee. Still there are some minor things you need to take care yourself
