Inspiration
The project was inspired by the observation that many businesses have to embark on long procedures to obtain basic insights from business performance data. This app aimed to make such a process shorter and more streamlined.
What it does
This app is an app within an app (explanation below). The app (operating through Streamlit) allows for business personnel (e.g., a data analyst) to quickly and easily create a preliminary version of a business performance report, based on cleaned business performance data. The app employs Streamlit to generate useful data visualisations, and Snowflake Arctic to obtain insights from the data and provide recommendations to the business. Additionally, the app allows for text translation, which may be beneficial if a business has non-English speaking stakeholders.
Challenges we ran into and how we built it
The app was originally only built in the Streamlit platform in Snowflake. However, certain desired features could not be displayed or operated, as per Snowflake policy (e.g., html text). To work around this, the main part of the app was written using the "normal" Streamlit platform, while the section of the app requiring the use of Snowflake Arctic was written in the Streamlit platform in Snowflake. The two apps were linked by using a button function (in "normal" Streamlit) that opened (a new web page) with the Arctic app in Snowflake.
Accomplishments that we're proud of
The app operated as expected and was capable of generating a preliminary business performance report in English. Additionally the app was used to translate the English report (into French).
What we learned
Apps that incorporate AI can be task or sector specific (e.g., the current app is targeted at businesses). Additionally, AI can assist one (or businesses) to gain a preliminary understanding of operational data.
What's next for Business Performance Report Generation App
This app may be developed to be more generic (i.e., applicable to any business with its own data format). The Arctic part of the report may be developed to simultaneously provide and compile all data insights into a format suitable for reporting. Additionally, the Arctic part may be improved to perform all the tasks of data receival and cleaning, data visualisation and analysis, as well as provide actionable recommendations.
Log in or sign up for Devpost to join the conversation.