AI Chatbot

Successful chatbots for use within your team don’t just need the best AI, they also need API access to all other applications. This allows the chatbot to fulfil user requests by looking up data, visualize it or start requests on their behalf.

Chatbot

The Assistant is like a companion the user can ask questions about their personal data, such as issues, emails or files. But they can also ask the Assistant about policies, easily start requests or find files across a range of connected apps.

The chatbot itself is a component of the Digital Assistant web app, but it can also be embedded into chat platforms or Intranets.

Proprietary NLP engine

Designed for business use cases, where there are many synonyms and acronyms for a comparatively large amount of intents, our NLP engine is easy to setup, transparent to admins and optimized to work with small amounts of training data.

Predictive Typing

When the user starts typing, the AI suggests questions or commands that it knows answers to. This gives the user an instant idea of the kind of knowledge the Assistant holds, but also serves as a hint to what the AI further expects from the request, e.g. the name of colleague or location, to turn the suggestion into a valid search command.

Knowledge Answers

The chatbot has two ways to answer questions: either by directly drawing live data from the backend system pertaining to the user’s intent. Or administrators can add large amounts of knowledge to the Assistant, which it will send to the user if a match is found in the form of a Q&A style Card. These Q&A Cards are useful for more generialized questions, such as policies, guest WiFi code, FAQs etc.

Recommended Content

With the use of Machine Learning algorithms, Digital Assistant will builds profiles from usage analytics to suggest content to users that are similar to profiles the algorithm created. Admins can modulate and add their own recommendations as part of the ELK stack.

AI Training

The accuracy of the AI Engine can be continually improved with the use of Utterances that are fed back into the training data, based on collected analytics reports.

Multi-lingual

The NLU processor is capable of being trained in multiple languages simultaneously, enabling it to even understand inputs delivered in a language different to the device language.

Embed answers into Search

Organizations that already use an Enterprise search or an Intranet may want the Assistant to deliver Google-style results directly on the search page. This is possible by adding the suitable Channel to Digital Assistant.

Some features are not available on all plans, please see the Pricing page for more info.