Skip to content
mwpreston dot net

mwpreston dot net

Primary Menu
  • Home
  • Certification
    • VMCE
    • VCP 5
      • Study Guide
      • Practice Exams
    • 8 weeks of #VCAP
  • Events
    • Storage Field Day 13
    • VeeamON 2017
    • Tech Field Day 12
    • VeeamON 2015
    • Virtualization Field Day 5
    • Virtualization Field Day 4
    • VeeamON 2014
  • Shares
  • Aboot Me eh?
  • Contact Me
  • Home
  • 2021
  • October
  • 20
  • Automating the creation of an AWS Lex and Lambda chatbots with Python

Automating the creation of an AWS Lex and Lambda chatbots with Python

Mike Preston October 20, 2021 7 min read

Part of my role at Rubrik is to educate and help customers and technologists who want to develop scripts and solutions utilizing APIs. One of my favorite ways to do that is through the creation of a chatbot, at Rubrik, we use Roxie, Rubrik’s intelligent personal assistant. Roxie is built utilizing two AWS services; Lex and Lambda. Lex handles the conversational pieces, while Lambda goes out and fulfills the question (or, you know, gets the answer). While it’s most definitely a great exercise to go through the process of creating Roxie manually (I promise, you’ll learn lots about AWS and consuming APIs in general), it’s also not something I enjoy doing over, and over, and over, and over (you get the picture). So with that said, during my last little development stint with Roxie I decided to figure out how to automate the complete creation of both the Lex Bot itself and the Lambda function! My weapon of choice, Python – the results, well, I think they are pretty good! So with that said, let me walk through the process of automating the creation of a Lex and Lambda chatbot with Python.

Prerequisites

As with anything tech/developer-related, there are a number of prerequisites that must be met in order to successfully follow along here, some obvious, some are not…

  • Python – yeah, you’ll need python installed on your workstation
  • AWS Account – pretty obvious, but you will need an AWS account to play with
  • AWS CLI – Having the AWS CLI installed and configured just makes things easier, that way we can just grab access and secret keys for the python code directly from your credentials file.
  • Boto3 SDK – We will be interacting with a number of AWS services via Python, so boto3 is a must! (pip install boto3)
  • AWS Lambda Role – An IAM role within AWS to assign to the Lambda function
  • AWS Lex Role – An IAM role within AWS to allow the Lex bot to run under. This is autocreated when you manually create a bot, but not when you programatically do it. You are best to simply go and manually create a test bot, get the Role ARN, then delete the bot 🙂

With these in place, let’s just jump right into the code!

Step 1 – Create the Lambda Function

Sometimes it’s just easier to see all the code and then break it down further afterward.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
import boto3
import json
import sys
 
config = {
    "aws_lambda_function_name": "roxie",
    "aws_region": "us-west-2",
    "aws_lambda_subnet_ids": "subnet-0bd3e627ba5bdb539,subnet-07a7efd2e2b097d64",
    "aws_lambda_security_group_ids": "sg-0cf9d4ed6393e46e7",
    "aws_lambda_runtime": "python3.7",
    "aws_lambda_role_arn": "arn:aws:iam::1233456789:role/service-role/roxie_lambda",
    "aws_lambda_environment_var_1": "10.8.107.34",
    "aws_lambda_environment_var_2": "BIG_LONG_AUTH_TOKEN"
}
 
from botocore.config import Configboto_config = Config(
    region_name = config['aws_region']
)