AutoNation Appointment Bot

Inspiration

The AWS Lex workshop and Pytorch Video

What it does

Almost lets you book an appointment and can let you select the day, the time, the reason, and the make of the vehicle

How we built it

Using the AWS Lex workshop example as a base for the bot, we went ahead and edited it to the best of our ability to make it book appointments. The plan originally was to split off and code in certain sections of the json files provided in the challenge but we were running out of time and decided to try and work on one part of near the end of the submission period.

Challenges we ran into

Getting Lex to recognize the lambda code, getting lex to give the correct output, getting lambda to pull the correct slots from lex, and lastly getting the lambda to be able to correctly detect the session attributes (which is still an issue but it will be fixed)

Accomplishments that we're proud of

Getting to be part of the hackathon as a whole. Having the bot even function giving us the correct information. Being able to say we came out of the experience having learned alot more than we anticipated.

What we learned

Learned about the AWS Lex platform as well as the Pytorch platform. We learned what are the foundational pillars of a machine learning algorithm and how to make one ourselves. Lastly, we also learned to be more humble in our expectations for what will be the final project.

What's next for AutoNation Appointment Bot

Finishing the appointment section and moving on to making it fully capable of reading a json file with a set of appointments, repair orders, customers, inventory, and much more.

Built With

• aws-lambda
• aws-lex
• python