Discourse.ai + Amazon

Discourse.ai is an Amazon Partner Network (APN) Registered Partner


As Amazon Connect customers move to launch Amazon Lex as their bot building platform, they look to Discourse.ai to help them automate the up-front processes required to set bot strategy and to reduce the time to design and train bots.

Why? Because accelerating bot deployments means speeding time to value for the business. 

Companies will save $11B* annually based on $2.2B** in annual spend on bot automation, however bot development is time-consuming and costly, and bots often fail to serve the needs of the end customer due to poor quality and the lack of volume of training data.


The Challenge? Building bots is a slow process… and you need to move fast.


Defining your bot strategy can take upwards of 6 months of intensive manual data labeling and analysis to determine the likely candidates and priorities for automation.

And the bot intent deployment process can take 6-8 weeks for design, development and training. And that doesn’t include tuning time…

Discourse.ai speeds time to value by as much as 75% by automating much of the manual effort in bot building.

Discourse.ai has integrated its Cognition Platform with Amazon Lex so your business analysts and data scientists can discover rich insights from your existing voice, live chat and chatbot conversations. And your bot developers can use your rich conversation data to train your bots.


Discourse.ai Cognition + Amazon Lex. How it Works.

Amazon 3 Step wConnect.png

Features you’ll enjoy with our Amazon Lex integration


Automated Conversation Labeling / Annotation

Amazon Connect voice call transcriptions, as well as the customer’s live chat and chatbot conversations can be loaded into Cognition for automatic labeling to enable visualization and analysis.

Data discovery of conversation patterns

Visualize conversation insights from bot and agent conversations to quickly identify the right intents and flows to implement or to identify improvements that should be made to enhance performance.

Export Amazon lex bot training data from the cognition Conversation Graph

Easily export intents, slots, entities and flows with sample data to load right into the Amazon Lex platform.

aws lambda integration with cognition api

Capture information from Amazon Lex bots to make bot behaviors and flows immediately visible in the Cognition Conversation Insights dashboard.

flexible deployment options

Discourse.ai offers Cognition as a SaaS or Premises solution globally, so we go where your Amazon Lex bots go…


Speed your time to value with Amazon Lex bot deployments.


Automatic Data Labeling will reduce the time required to prepare conversation data for analysis by as much as 75%, and reduces costs from what is typically $2 - $5 to $.25 - $1.25 per conversation labeled.

Visualization of ALL of  your customer conversations provides insights in to customer intents and opportunities for automation, speeding time to set bot strategy and defining bot designs by 50-75%.

Extracting a high volume of customer utterances to automatically load into your Amazon Lex bot saves your team from the manual effort required to create bot content, saving another 50% - 75% labor cost and time per intent.