RoboExtract
- Denny Alfath
- Aug 31, 2022
- 2 min read
RoboExtract is a web based application that help users to extract data from any type of documents with the assistance of Artificial Intelligence.
Overview
The Problem
Staffs who work in logistics industry need to extract data from a scanned document by typing it manually into an Excel file. This became their daily task and they need a way to save time from this task.
The Goal
Save time that needed to extract the data and spend more time to high-priority tasks.
The Outcome
The main deliverable of this project is a fully functioning web based application where users can use Robot to extract their documents automatically.
Process
Research and Discovery
To start this project, i did a research to discover opportunity with these questions leading the process:
1. What are the user and business needs?
2. What technology capacities that the company has?
My limitation in the research process is I can't directly reach the existing users to test the product and investigate deeper about their pain points. To handle this limitation, i did an interview with one of our Customer Service staff who always help the users using the product. Because they spent the most time with the users.
Interview Findings
• Lack of clear navigation that confused the user after did a certain process.
• The words used are not user friendly and it is hard to understand.
User Persona
Solutions
After re-discuss the problems found to the Customer Service team, we came up with these solutions:
1. Provide a sidebar navigation to help user navigate through the whole feature of RoboExtract.
2. Provide a user-friendly words to help user understand how to run the data extraction in RoboExtract.
User Flow
Wireframe
These are the base concepts for overall page. The extraction process will start from the labelling part where user labels the tags so the machine can be trained to learn the pattern of the data in Train Label process. And finally using a model generated from the training process, user can do the data extraction.
User Interface
Colors and Typography
Primary
Accent
Comments