Designing a reusable
microservice
A scalable intelligent task management microservice built for an enterprise application—enabling associates, vendors, and managers to organize, prioritize, and complete critical actions on time in one unified place.
Year:
2026
Timeline:
12 weeks
Overview
Vendor Gateway is a B2B platform used daily by vendors and internal associates to manage business operations. This platform is used by 20+ teams who have integrated their applications as MFEs.
Each application's users were maintaining the tasks on spreadsheets, and there was no centralized system creating operational inefficiency, missed deadlines, and overcommunication.
The vision was to create a centralized and personalized task manager microservice that could aggregate tasks from multiple applications into a single, scalable experience while remaining flexible enough to support different teams' needs.
Why this case study matters?
This isn't a case study about picking the right layout or just a design experience. It's about how research and testing help prioritize and deliver the right features.
My Role
As a Product Designer, I worked on across the full design and deliver phase.
Led requirement-gathering workshop and sessions with PMs across multiple teams which surfaced JTBD, feature priorities, current nuances and tasking workflows.
Designed and rolled out the vendor questionnaire to understand tasking behavior and needs.
Explored the concept directions and led the evaluation process.
Collaborated with multiple designers to meet every teams requirement.
Ran usability test to directly reshape the product.
How it started
Based on the qualitative findings from a user research with Vendors, we found out that communication was a real challenge with approximately 40% users wanting a way to show the tasks they need to complete and be more aware when any deadline is coming.
Goals
Reduce the time spent finding and actioning tasks
Reduce dependency on support teams for task-related issues
Increase visibility to plan and prioritize
Establish a scalable framework for all consuming teams
Challenges
Key challenges that we identified even before starting with the research and workshops.
01
Scattered Tasks
Tasks were scattered across multiple applications with no way to keep a track in a centralized place
02
Manual Tracking
Associates working with multiple vendors had to manually track tasks for instance, whose contract is getting over and when
03
Missed Deadlines
No prior communication or reminders to vendors, leading to frustration, especially for compliance issues.
04
No ownership
Tasks remained in excels with no ownerships and audit trails, leading to business critical issues.
A lack of transparency in status tracking led to vendor frustration, contributing to higher churn and lower trust levels, as indicated by low LTI scores. Associates also spent considerable time managing status inquiries through email, impacting overall operational efficiency.
Step 1: Vendor Questionnaire
Understanding the real needs before designing
Rolled out a survey to understand real tasking behaviour: how many tasks, from which application, how often deadlines were missed, whether any collaboration happened, and what notifications preference looked like.
Step 2: Workshop
Cross-team workshop: Understanding workflows
Conducted a workshop with 10+ PMs from different teams to understand major workflows/service blueprint, what is hard today, JTBD and feature prioritization.
What I wanted to understand
No. of tasks and frequency for vendors and associates
Are they currently or in the future generated automatically by the system or created manually by the user?
Are the tasks enough to have its own dedicated tab?
Types of tasks— Forms/documents/view-only, etc.?
How will it get assigned to the right user?
How direction got changed
Numbers don't lie.
01
Associates became our priority
Vendors were actually dealing with 5-6 tasks per month, whereas associates were handling 100-7k tasks based on persona and teams in a month.
02
Vendors don't need a separate tab for tasks
Vendors wanted more communication and reminders on these tasks rather than a dedicated place to complete them.
03
MFE and Deep linking integrations for teams
Centralizing data from multiple teams posed a tech challenge due to the high data volume. Additionally, existing flows had to be leveraged to avoid extra rework.
Step 3: Define Phase
Design for not just an application but a reusable and consistent framework
I mapped the user journey for both vendors and associates, proposing to multiple designers and PMs to get buy-in before starting.
Went through the content to find themes and patterns that can be reused in the cards.
Step 4: Design Phase
Designed a customizable card to meet all teams need
The card comprised of 4 sections—
Badges to show category, sub-category, and status.
Primary Content Section consisting of vendor name and/or number for associates and task heading for vendors.
Metadata section with Task ID, Due Date, and 2 placeholder content.
Assignee section to show if the task is unassigned or to whom it is assigned.
We choose the Kanban Board design based on previous data
Because the design was already implemented in a different team for a different use case, the vendors found it intuitive and easy to use from their research.
Testing
Synthetic A/B Testing (Chatgpt)—
Conducted an A/B test on Saved Filters to understand user preferences and behavior when saving, renaming, and deleting filters used in their daily workflows using Chatgpt.
8 Associates/Managers Usability Testing—
Planned and led usability testing sessions to validate key workflows, including task self-assignment and task discovery, while capturing qualitative feedback on design effectiveness and ease of use.
Insight
Solution/Feature
Trade-Off
Why
User testing revealed that associates struggled to quickly locate their assigned tasks and were not happy with them being filtered from the filter panel. Many users also expressed interest in viewing tasks assigned to others, finding the additional information overwhelming and distracting.
Prioritized the "My Tasks" experience over email and notification capabilities in the initial release.
For associates, managing tasks was a core part of their daily workflow, making immediate task visibility more valuable than notification-based reminders. Unlike vendors, associates were already expected to log into the system regularly, reducing the need for proactive communication features in the first phase.
Managers found assigning tasks individually to be time-consuming and inefficient, particularly when they needed to allocate large volumes of tasks to a single associate.
Introduced a bulk assignment feature, enabling managers to assign thousands of tasks to a user in a single action, significantly reducing effort and improving operational efficiency.
Deferred the Saved Filters functionality and instead focused on capturing usage metrics and user feedback after launch to validate its importance.
Introduced the tool for only 2 teams, which had 10-15 tasks a day until the functionality was delivered.
Assigning tasks one at a time created unnecessary friction and risked reducing adoption of the tool. In contrast, filtering remained manageable with a relatively small set of available filters, while bulk assignment addressed a critical, high-frequency workflow for managers handling large task volumes.
Associates frequently wanted to filter and locate tasks using additional metadata fields. Existing filters were insufficient for the variety of information users relied on to find relevant tasks.
Expanded the search capability to support free-text, elastic search across task metadata. Previously, search was limited to VBU name and number, but the enhanced search enabled users to find tasks using a broader range of attributes.
-
Adding dedicated filters for metadata owned by other teams presented significant technical challenges. Dependencies on cross-team data integrations and MFE implementations were also slowing development. Enhancing search provided a faster, more scalable way to meet user needs without introducing additional filter complexity or integration overhead.
Final Design and Functionalities
Choose a previously designed option over kanban board—
Bulk assignment was more effective for handling large volumes of tasks and avoided the usability issues of a kanban board, where most unassigned tasks were concentrated in the “To Do” column. In that setup, selecting multiple cards would require excessive scrolling, making the process inefficient.
This approach also ensured:
Better visibility of tasks even after status-based sorting
A more mobile-friendly and responsive experience
Easier scanning and selection of tasks
Scalability across different teams and workflows
Impact
Data we got after the release.
6/10
teams integrated
20%
reduction in tickets
25k+
tasks processed
82%
tasks completed before SLA
Metrics Defined
Metrics we defined that need to be measured in different categories.
Adoption Metrics
Active Users (Daily, Weekly, Monthly)
User Retention Rate
Unique Users
Business Metrics
Task Completion rate (Average Lifecycle)
Overdue Task Rate
Support Ticket Reduction
Productivity Metrics
Search success rate
Bulk assignment adoption
Single vs bulk adoption
Pattern in filter usage
Notifications click-through rate
Reassignment rate
Abandonment rate (Search, assign, reassign)
+
+
+
View other
Case Studies
Let's Talk
Curious what we could build together? Let’s turn bold ideas into meaningful, high-impact experiences. Drop me an email at khuwalneha@gmail.com to learn more — or just to chat over a cup of coffee.
+
+
+