I attended AWS Week at the Loft in San Francisco where I first heard about Lex. I have developed different financial calculators in the past. And these calculators are often dull resulting in poor user experience. I was inspired by the power of Lex with Lambda to implement a Chatbot for one of these calculators.
What it does
RetirementBot asks basic questions about you , your employment and other financial status and using that calculates the amount you need to have in savings for your retirement. If the user doesn't have an answer, the bot makes intermediate calculations and get confirmations from the user.
How I built it
I did a proof of concept with a template and then a custom bot. Once I got comfortable with that, I create the framework for the RetirementBot. NorthCapital had the retirement calculation in an Excel sheet and we implemented this in the Chatbot.
Challenges I ran into
Negative conditions are not directly handled. Facebook page deployment had its own challenges.
Accomplishments that I'm proud of
We now have a functioning chatbot and can leverage the power of Facebook's distribution to get attention.
What I learned
Amazon's cloud offering is mature but documentation for various features are not clear and hence you need to dive deep and keep focus. It is a robust platform and things will fall in line and start working.
What's next for RetirementBot
I plan to enhance some of the functions to make it more interactive. Also, we plan to create a Slack channel with the same code, deploy behind an SMS number and also as an Alexa skill. All of these will add to the product suite for NorthCapital.