How NLX works

Explore the power and versatility of NLX with your existing tech stack

Whether deploying a conversational AI application through a voice or chat channel, offering enhanced multimodal interactions, or incorporating your knowledge base, the NLX platform allows you to construct and deploy the most robust applications with its advanced conversational AI builder.

What the NLX platform handles

NLX integration architecture

• Applications: Stores and manages your conversational AI applications that can be organized by line of business, user group, production environment, channel type, or whatever makes sense for you
• Conversation management: The design and construction of conversation workflows that automate tasks and deliver responses. This includes training data to help educate AI models on which flow to trigger when a user provides a query (utterance)
• Integrations management: Exposes the endpoints for you to integrate AI model services (NLP, LLM), communication channels (web, WhatsApp, Amazon Connect), as well as your own systems of record to retrieve or send data during conversations

• Knowledge management: Exposes the endpoint for you to integrate your knowledge base and allows you to construct and manage all LLM prompts when retrieving data

• i18n management: Allows you to specify the languages your conversational AI application supports and manage and store all translations (manual and auto-translation options available)
• Insights & analytics: Real-time analytics with dozen of stored metrics to yield thousands of data points in custom dashboards. In-flow analytics also available to demonstrate traffic data happening within each conversation workflow
• Alerts: Easily monitor key metrics by setting up custom alerts to notify you of critical activity

• Conversation history: Stores and provides complete transcripts of all conversations occurring between users and your conversational AI applications

• Roles & permissions: Provides advanced access control for you and your team within the platform

---

What the NLX NLU does

To help provide access to what your conversational AI application needs when servicing users, NLX's NLU manages the interactions happening between your frontend interface, cloud-based or on-premise backend systems, and preferred NLP, LLM, and CCaaS services.

The NLX NLU works by

• sending and receiving data from a frontend channel (where users converse with your AI application)

• sending or retrieving data from a backend system through a custom API integration

• storing all logic, messaging, and AI prompts that make up your AI application into a single package (or build) for deployment

As all conversations are comprised of turns between participants, the NLX NLU also keeps track of turns taken in an exchange between your users and your conversational AI application (conversation state) . This includes where in the conversation a participant is (turn).

By managing and storing conversation state, the NLX NLU not only keeps the conversation operational, but it relays this valuable information when providing full conversation transcripts, troubleshooting conversations, and detailing hundreds of data points for advanced analytics.