aws-solutions-library-samples/qnabot-on-aws

What is it?

Key features

• 01Answers questions over chat, voice, SMS and Amazon Alexa through Amazon Lex.
• 02Content Designer UI for authoring and managing question and answer entries.
• 03Semantic matching with text embeddings and generative answers via Amazon Bedrock.
• 04Fallback to Amazon Kendra or a Bedrock Knowledge Base when no answer matches.
• 05Optional guardrails to scan and mask inputs and responses.
• 06Usage logging and dashboards for unanswered questions and user feedback.

Use cases

• Deploying a customer support FAQ bot that answers common questions across web chat and phone.
• Reducing call center wait times by automating routine support questions through Amazon Connect.
• Building a question-answering assistant grounded in a company's own documents via a Bedrock Knowledge Base.

Quick start

1. 1

   Install Node.js, npm, git LFS, Python 3.14+, Poetry and the AWS CLI, then clone the repo

2. 2

   Install QnABot's Node modules from the source directory

   npm install
3. 3

   Generate config.json, then set region, profile, namespace and devEmail

   npm run config
4. 4

   Create the bootstrap S3 bucket, then deploy the stack (run bootstrap first, then up)

   npm run up

Things to know

• !Custom deployment requires a Linux environment, with Amazon Linux 2 as the tested platform.
• !Setup needs several toolchains (Node 24+, Python 3.14+, Poetry, git LFS) and a deployed AWS account, which will incur AWS costs.
• !Deployment depends on Amazon Bedrock, OpenSearch and other AWS services, so it is not a self-contained local install.

In collections

Topics

FAQ

What is qnabot-on-aws?

A multi-channel, multi-language chatbot on AWS that answers customer questions over chat, voice, SMS and Amazon Alexa, managed through a Content Designer UI. QnABot on AWS is an AWS Solutions implementation that deploys a conversational question-and-answer chatbot into your own AWS account. Questions and answers are stored in Amazon OpenSearch Service and managed through a Content Designer web UI, and the bot is built on Amazon Lex so it can be reached from a web client, Amazon Connect, SMS, voice and Amazon Alexa.

Is qnabot-on-aws free to use? What license does it have?

qnabot-on-aws is open source under the Apache-2.0 license. Check the license terms before using it in commercial products.

Is qnabot-on-aws still maintained?

Updated within 6 months. The last commit to qnabot-on-aws was on Sep 9, 2026.

What can I use qnabot-on-aws for?

Deploying a customer support FAQ bot that answers common questions across web chat and phone. Reducing call center wait times by automating routine support questions through Amazon Connect. Building a question-answering assistant grounded in a company's own documents via a Bedrock Knowledge Base.

How do I install qnabot-on-aws?

Install Node.js, npm, git LFS, Python 3.14+, Poetry and the AWS CLI, then clone the repo Install QnABot's Node modules from the source directory: npm install Generate config.json, then set region, profile, namespace and devEmail: npm run config Create the bootstrap S3 bucket, then deploy the stack (run bootstrap first, then up): npm run up

Where can I get qnabot-on-aws?

The source code and installation instructions are on GitHub: https://github.com/aws-solutions-library-samples/qnabot-on-aws