Natural Language Platform

1The AI Feedback Loop

The AI Feedback Loop consists of four layers:

Natural Language Understanding (NLU), a subset of Natural Language Processing (NLP), focuses on the semantic understanding and underlying intent of words. NLU provides value by performing critical operations on text data. Using BERT-based techniques, NLU enables AI to automatically extract information from any corpus. They also perform semantic modeling to understand language variations and to drive query resiliency. NLU also employs techniques for determining the intent of natural language communication as well as techniques for explaining search, discovery, and briefings so that users can tune the system.

No-Code Models enable humans to input enterprise knowledge into the system. In partnership with AI, humans can use these models for extraction tasks, semantic model evolutions, and intent scenarios.

Natural Language Services are used for system diagnosis and no-code tuning.

User Experience focuses on the overall user experience and is based on live testing, benchmarking, curation tools, and diagnostics. It lets humans to improve the accuracy of AI models using explainability techniques and no-code tuning.

2AI Augmenting Humans

The hallmark of the natural language-enabled enterprise is the immediate value derived from interacting with text data using natural language to mine, explore, and gain insights. AI Augmentation consists of three layers:

Natural Language Services Layer

Cognitive or intelligent search understands the underlying intent of searches and their specific vocabulary to deliver the most precise results with a question asked, question answered format.

Briefing enables users to rapidly skim through a large volume of information based on specific questions of interest. Users simply present their questions; the AI quickly parses the data and surfaces salient information. This enables users to gain insights from a much broader amount of information than they could read themselves.

Discovery allows users to find unanticipated or associated information related to their search. Discovery inspires new paths of investigation and hidden insights.

Self-Diagnostics Layer
‍The self-diagnostics layer enables the system to capture and analyze the behavior of the users to either self-correct or to help humans improve system performance. Both enable continual improvement.

Application Layer
‍The application layer ties together the individual services for a comprehensive collaborative experience between AI and humans. This layer also reports human behavior back to the self-diagnostic system for additional performance improvement.

Natural Language
Platform

Kyndi Natural Language Platform

Natural Language Enablement

The Kyndi Platform Is Built Upon a Collaborative Paradigm

Unleashing the Power of Human Intelligence