TELUS Digital - AI Platform
TELUS Digital (formerly TELUS International) designs, builds and delivers innovative solutions to enhance the customer experience (CX) for global and disruptive brands.
At Telus Digital Experience, I was part of the AI Data Solutions team, where I played a key role in developing the AI Platform.
The AI platform provides services for internal teams on areas such as sourcing, productivity, evaluation, etc. It also handles a community of people who perform tasks such as translation, annotation, transcription, etc.
My primary responsibility was to enhance this platform to support Telus’s global network of contributors and internal teams in overseeing, managing, and executing customer projects.
At Telus Digital Experience, I was part of the AI Data Solutions team, where I played a key role in developing the AI Platform.
The AI platform provides services for internal teams on areas such as sourcing, productivity, evaluation, etc. It also handles a community of people who perform tasks such as translation, annotation, transcription, etc.
My primary responsibility was to enhance this platform to support Telus’s global network of contributors and internal teams in overseeing, managing, and executing customer projects.
Date:
2022 - 2025
location:
Tampere, Finalnd
My Role:
UI/UX design
Wireframing
Prototyping
User flows
Design system
Visual design
Case Study : End to End project
Tools used:
Figma
Illustrator
Google tools
Photoshop
Context
The aim of this project was to simplify, unify, and automate the end-to-end production processes of internal teams while supporting them by deploying new functionalities, migrating their projects, and integrating their processes into the AI Platform.
This initiative was urgent because the primary tool used by the production teams was outdated and scheduled to go out of service within a few months.
This initiative was urgent because the primary tool used by the production teams was outdated and scheduled to go out of service within a few months.
Primary Research and key findings
We began this project by conducting interviews and surveys to identify key pain points faced by users and analyze the current processes.
Key Findings:
While there were no significant differences in the baseline processes between teams (our target users were in two different teams located in Finland and Ireland), there was no standardized process in place.
The existing process relied heavily on manual workflows and Excel trackers, making it tedious and prone to human error.
Independent and isolated self-solutioning resulted in the use of numerous duplicate tools to address the same business needs.
Wireframing
I began by reviewing the research findings with a Product Manager to discuss how we could support users' needs and ensure a smooth transition from their current process to the AI Platform.
Given the tight timeline for this initiative, we prioritized addressing the most pressing issues first.
Since the primary tool for tracking the headcount of contributors working on the project was about to go out of service, we focused on developing an MVP to address this need.
Given the tight timeline for this initiative, we prioritized addressing the most pressing issues first.
Since the primary tool for tracking the headcount of contributors working on the project was about to go out of service, we focused on developing an MVP to address this need.
Prioritised pain points:
There was no real-time data available on the number of contributors actively working on a project.
The teams relied on a Google Sheet to track contributors.
If a contributor went on vacation or resigned from a project, they had to manually cross-check emails from contributors with the Google Sheet, which contained contributor lists, to identify absences or resignations and then update the records manually.
This process was tedious, time-consuming, and prone to errors.
If a contributor went on vacation or resigned from a project, they had to manually cross-check emails from contributors with the Google Sheet, which contained contributor lists, to identify absences or resignations and then update the records manually.
This process was tedious, time-consuming, and prone to errors.
Contributors had to either use an outdated tool to submit a form or email the production team directly to request absences or resignations.
There was no centralized system to store all the information in one place; the data was scattered across various tools and Google Sheets.
Additionally, the process lacked automation, consuming valuable time for production workers, whose primary role was to oversee projects and evaluate the work contributed by each individual.
Some projects had over 10,000 contributors, making this inefficiency even more impactful.
Additionally, the process lacked automation, consuming valuable time for production workers, whose primary role was to oversee projects and evaluate the work contributed by each individual.
Some projects had over 10,000 contributors, making this inefficiency even more impactful.
MVP wireframe creation
To address these needs, I began defining the user flow and designing MVPs for two key functionalities:
Production Tracker:
An automated, real-time list of contributors for each project, allowing production team users to check the current status of contributors.
My Availability:
A centralized system where contributors can submit absence requests and resignations, which automatically update the Production Tracker.
Production Tracker
We began developing the Production Tracker MVP by identifying the key information needed by the production team to successfully initiate and complete a project. To do this, we consulted with production team members working on various client projects.
By analyzing their responses, we decided to introduce several automated status types:
By analyzing their responses, we decided to introduce several automated status types:
Lifecycle Status:
Indicates whether a contributor is working on a different project or available for a new assignment.
Profile Status:
Reflects whether a contributor is actively working or absent.
Project Status:
Displays the contributor’s current situation within the project.
The names of these statuses were based on terms familiar to the production teams.
Production Tracker MVP example
My Availability
We began developing My Availability as a platform where contributors could submit absence requests or resignations for the projects they were working on.
The submitted absences or resignations would automatically be reflected in the Production Tracker, allowing production team members to view the number of active contributors in a project in real time.
The submitted absences or resignations would automatically be reflected in the Production Tracker, allowing production team members to view the number of active contributors in a project in real time.
My Availability MVP example
After multiple iterations and collaborative feedback sessions with product managers, production team members, and the development team, I refined the design by incorporating feedback and ensuring alignment with product goals. Once the key aspects were finalized, I proceeded with prototyping to create a more interactive and tangible representation of the final product.
Prototyping
I created prototypes based on the feedback I received, ensuring they met the business requirements.
The prototype was designed to cover all basic scenarios and user flows while remaining flexible and scalable for future enhancements.
Once the prototype was complete, I shared it with developers and product managers.
Once the prototype was complete, I shared it with developers and product managers.
Production Tracker prototype example
My Availability prototype example
Testing
Testing was a critical part of the design process. I shared the prototype with internal users, alongside the Product Manager, to validate the usability and effectiveness of the functionalities.
The feedback was overwhelmingly positive, with users expressing that the new features would make their workflow significantly easier and less tedious.
After implementing the necessary refinements, I handed the design over to the QA team for in-depth internal usability testing.
Next step and what I learnt
After we implemented the first MVP into the AI platform, other production teams began expressing interest in migrating their projects to the platform as well.
With the increase in requests, we started adding additional functionalities to My Availability and the Production Tracker.
As we received more functionality requests, it became evident that the scope of the project was expanding, requiring adjustments to the existing features to accommodate the added demands.
This highlighted that the initial solution could have been designed to be more flexible. For example, one key functionality we later added was the ability for users to create customizable labels and columns.
The next step for the project is to make the tracker more flexible and automated while integrating other aspects of the production team's workflow. The goal is to create a platform where the team can manage all areas of their work seamlessly.
One of the features currently in development is functionality that allows production team members to view all evaluation results of contributors’ work, as well as enhancements to the contributor side of the platform.
With the increase in requests, we started adding additional functionalities to My Availability and the Production Tracker.
As we received more functionality requests, it became evident that the scope of the project was expanding, requiring adjustments to the existing features to accommodate the added demands.
This highlighted that the initial solution could have been designed to be more flexible. For example, one key functionality we later added was the ability for users to create customizable labels and columns.
The next step for the project is to make the tracker more flexible and automated while integrating other aspects of the production team's workflow. The goal is to create a platform where the team can manage all areas of their work seamlessly.
One of the features currently in development is functionality that allows production team members to view all evaluation results of contributors’ work, as well as enhancements to the contributor side of the platform.
Seibun Design System
In addition to designing new features, I contributed to updating and expanding the design system, Seibun UI, ensuring consistency and scalability across the platform.
Seibun UI
Example of Seibun UI components
Context
The AI platform went through multiple transitions and rebranding phases. We noticed that some components came from legacy or third-party libraries. To ensure consistency and provide a streamlined experience across the platform, we needed to update the design system.
Reference and structure
We conducted benchmarking using established design systems. After several meetings with the developer to go through the findings, we decided to mainly use Carbon Design System (IBM), Dell Design System, and Material Design (Google) as references.
To organize the components, we decided to categorize them into six groups: Style Tokens, Input Controls, Data Display, Information Components, Navigation, and Layouts.
To organize the components, we decided to categorize them into six groups: Style Tokens, Input Controls, Data Display, Information Components, Navigation, and Layouts.
Adding Components
I added or updated components across various categories whenever I saw the need for them while developing new features or flows. The usual process involved first creating a draft component along with examples of its behavior. Then, we reviewed them with developers by organizing collaborative sessions to assess feasibility and potential redundancy. After receiving feedback from fellow designers and developers, I revised the component if needed.
Holding sessions with developers ensured clear communication and streamlined the implementation of the design system.
Holding sessions with developers ensured clear communication and streamlined the implementation of the design system.
Example of a component and a component behavior document
What I learnt
This experience deepened my understanding of design systems and their role in ensuring visual coherence, scalability, and alignment with brand identity and user expectations.
It also strengthened my collaboration with cross-functional teams, where effective communication with developers fostered a smooth workflow from drafting to implementation. This foundation not only streamlined our process but also created opportunities to refine and elevate the design system further.
It also strengthened my collaboration with cross-functional teams, where effective communication with developers fostered a smooth workflow from drafting to implementation. This foundation not only streamlined our process but also created opportunities to refine and elevate the design system further.