How I made data onboarding 4X faster 🚀
Command Palette
Search for a command to run...
Comments
Join the discussionNo comments yet. Be the first to comment.
Search for a command to run...
No comments yet. Be the first to comment.
While I was working at Assetowl, a property management solutions company in Australia, I faced and solved a big problem that greatly improved how we brought data into our system.
When acquiring new customers we had to import their past data as a part of the onboarding process. Customers gave us inspection data in different formats, including PDF, Word, and Excel/CSV files for 100s properties they manage. In the onboarding process, this acts as a bottleneck because ~8 data entry operators had to do this manually. The variety of these data formats made the process of adding this data to our system slow and more likely to have mistakes. This was a huge challenge to make sales.
My main responsibility was to oversee data onboarding. I had to witness the difficulties and hurdles during this process. Top management had plans to implement a system to facilitate data onboarding but since Assetowl was a startup company top management did not have enough resources to invest in the onboarding system.
To tackle this challenge, I developed Python scripts, leveraging the capabilities of the Openpyxl library. This programming approach was aimed at converting data from CSV files into JSON files. JSON, or JavaScript Object Notation.
This technical solution significantly minimized the need for manual data entry. Automating the conversion process, not only saved time but also greatly reduced the possibility of errors. The code for these scripts is accessible on GitHub at this link: PDF-to-JSON.
Now I had JSON files ready to import into the system but there was an issue. Due to the inconsistency of customer-provided files, the JSON files were not 100% accurate. Even though JSON is humanly readable it was really hard to edit the JSON files.
As a solution, I designed and developed a quality assurance web application using React, Typescript, and Redux. This application allowed users to upload above mentioned JSON files, make edits, and download the edited versions with ease, enhancing the overall data onboarding experience.
The source code is available on GitHub at this link: pcr-qa on GitHub. Additionally, a live demo of the application can be accessed at pcr-qa.netlify.app, providing a hands-on experience of its functionality.
With these solutions, data onboarding was 4X faster than manual data import. As a result of this increased efficiency, we were able to onboard more customers than ever before. The faster processing times allowed us to handle a larger volume of data without compromising on accuracy or quality. This was a significant step forward in meeting the growth of our clientele.
