Finexa - your private finance assistant. Phone apps to control finance were a great achievement when they were introduced. But today, it is a pretty obvious thing. In such case, how do we make it better. Think of it this way - You want to know your previous week transactions but do not want to browse through all those long list of transactions. Your mother cannot work around with the new app on her own but wants to check her funds constantly In comes Finexa. Finance details are a mere conversation now.
What it does
Finexa is an alexa skill that lets you check your account details and past transactions without having to pickup your phone. Want to check the balance? Just say “Hey Alexa, what’s my current balance”. Not restricted by that, it can also segregate the transactions based on their type and provide deeper insights into the expenses in temporal fashion.
How we built it
Finexa uses alexa skill development to extract transaction details. The information is extracted using a python REST API in flask hosted on Amazon AWS. It provides with all the transactions and their categorizations. Next, we used Alexa skill development kit to train alexa. We use node.js to design the amazon lambda function. Just add your account details once and alexa fetches the transaction details for you everytime.
Challenges we ran into
Developing the project had its own challenges. Even though API’s exist to extract transaction details, they are not easily available due to the high level security constraints for financial data. Thus, we created our own data using MongoDB instance on AWS. We created an API using python flask to make calls from Alexa. This also had its own challenges.We had security issues with MongoDB Atlas and EC2 due to firewall and compatibility issues.
Accomplishments that we're proud of
We have successfully made it possible to convert the tedious task of going through all the transaction details into a simple conversation with Alexa. Similar products(Alexa skills) exist in the industry but are limited to simple balance based inquiries, but unable to factor in categorizing and other possible ways of analyzing the transactions. We have successfully learnt Alexa Skills and AWS Lamda Function with our own MongoDB database.
What we learned
What we take back from this project is the vast bandwidth of application for alexa skills development. Alexa skill along with Lambda function provided the convenient platform to train alexa. AWS services provided all the resources to deploy the API’s for data as well as deploy the final application.
What's next for Finexa - Finance with Alexa - Advance Track
The outreach of finexa is unrestricted. Using the high-end API’s with high level security would provide actual data for user’s. It can be made into more of an industry-level alexa skill by creating user authentication, voice recognition, machine learning algorithms to incorporate suggestions based on user transactions, etc. The limit is your imagination!