Continuous monitoring

Analytics and Profiling

Chatlayer continuously monitors your chatbot. We analyse conversations, check how you configured your NLP and give you concrete advice on how to improve your bot.

NLP Analytics

On our NLP dashboard, you get an overview of how well your NLP model is performing. For all your intents we analyse the number of expressions, their diversity and how and why misclassifications happen.

Turn NLP improvement into a game

Each time you train the NLP you will see the result. And you can also follow how well you have been doing. This way you can clearly monitor and report your progress, but it also turns building a good NLP into a game.

Improve your bot

On the NLU Performance dashboard, you can click through to our improve module. Here you get a step by step interface that guides you with specific advice on how to improve your bot.

Quality monitoring and flow analytics

Evaluate the best flows and use data to improve the conversion rate of your flows and the quality of the user experience. Assign groups of users to a specific flow, and gain insights on their feedback and conversion rates. With our flow analytics tool, you can dig through all your conversations. You can see where your users dropped off, how the felt and which conversations are used most.

Sentiment analysis

Chatlayer also analyses the sentiment of your users. Are your user not happy or neutral but dissatisfied, then you know where you should improve your conversational design.

Monitor your users

Monitoring your current users allows you to get a feel for the overall popularity of your chatbot and its success. The recurring active users gives you an indication on how often users are coming back to your chatbot after the initial use.

Messages and sessions

The amount of sessions presents the concept
of engaged users. Session duration measures
average customer interaction and ensuring
customer engagement.
Messages per session, shows the amount of messages an average user sent to get the answers he was looking for.

You shouldn’t take it for granted when you could actually get proof with a demo.