VoiceGopher

Inspiration

Reporting capabilities have grown during last few years immensely. There are dozen of engines and languages that will help you to slice and dice, crunch and process your data as well bunch of visualization frameworks. What is missing is vocalization and audio control of reports and spreadsheets. This fact inspired us to create platform and skill for vocalization of spreadsheet queries, which will benefit large group of users. As an example, the skill will help visually impaired users to better approach and analyze the data.

What it does

You have to create an account in our website, link your web account with Alexa via simple alexa command upload you spreadsheets. Once spreadsheet is in S3, you can invoke simple statistical commands on spreadsheet, like find minimum value of column B, or Tell sum of Column 3.

How we built it

We used Nodejs in Lambda and Ruby for Web side. Inside of postgres we created Users table, where we insert User once you register in Website. after you register, you get confirmation code that you have to Tell Alexa to link your Alexa userid with Amazon userid.

Challenges we ran into

Properly linking accounts via confirmation codes

Accomplishments that we're proud of

Security and isolation of data for individual users
Parsing spreadsheets and allowing users to target columns either by number or their cell names in excel (A, B, C.. AA..)

What we learned

So far there is no way to update slot types dynamically. That would allow us to query spreadsheets by column headers.
Work with S3 from Lambda service on behalf of Alexa