Telegram bot on AWS Step Functions

Serverless Telegram bot made on 4 AWS Lambda chained by AWS Step Functions. All of this written on Serverless Framework using plugins.

This is an example on how to build a simple Telegram bot backend using AWS Lambda functions and AWS Dynamo DB to store messages and AWS Step Function to chain Lambdas.

Deploying

1. Setup AWS credentials
2. Install Serverless Framework
3. Install two plugins: serverless-step-functions and serverless-pseudo-parameters
4. Clone this repo
5. Deploy: sls deploy

Architecture

All messages from your Bot will come to the State Machine first.

State machine will invoke two Lambda function in parallel with same payload:

log - will just save payload to DynamoDB and receive - will parse the payload and check for the specific auth message.

receive will pass the result to the next state: AuthOrNot - which is not a Lambda function but just a decider. It will check the result of previous function and either invoke authorize or respond Lambda function.

authorize is aimed to send a callback to your website to authorize the user if needed. The respond function will just send a response to the chatbot.

Too complex?

Here is an example of super simple Telegram bot with only one Lamdba: https://github.com/Andrii-D/serverless-telegram-bot