Atomicwork, 2025
ATOMICWORK INTEGRATION- WORKDAY INTEGRATION
team
Product Manager
Design manager
Frontend devs
Backend devs
Marketing team
My role
UX flows
rule builder UI
edge cases
dev handoff
Duration
3 weeks
------o background and problem
Atomicwork building its Workday integration to automate app provisioning and deprovisioning based on employee data.
Workday is often the source of truth for employee details like department, location, and role. But syncing that data with app access wasn’t smooth. Admins had to rely on IT ops to grant/remove access manually, which wasn’t scalable.
Enterprise users were constantly switching between Workday and Atomicwork to complete basic HR tasks. This led to:
Workflow delays
Inconsistent or outdated information
Frustration among employees and HR teams
------o but, What is atomicwork?
Atomicwork offers an Agentic Service Management solution that combines AI agents, a modern service platform, and automation engines to simplify enterprise support and service operations.
------o User want one unified experience for workday
Set role-based access rules using Workday data
Preview who these rules applied to
Roll it out without relying on engineering or scripts
Atom interaction, Apply for reimbursement
AI controls solve this by giving teams the ability to fine-tune how the assistant works, define when it steps in, and decide who gets to manage those rules. Done right, this builds trust, reduces hesitation, and helps AI features land smoothly across different types of user
IT teams wanted to use Workday fields (e.g., department, location) as filters
Accuracy was key — any misstep could expose apps to the wrong people
Admins wanted to preview which users would be affected before enabling rules
FOCUS AREA
1. Map Workday fields cleanly into the UI
Admins should be able to define if-this-then-that logic to control access easily.
2. Show live user previews before confirming
Pulling and matching Workday fields should feel intuitive in the UI.
3. Safe, simple, and scalable
The experience should feel lightweight yet secure — and scale without friction.
Flow 1: Reimnursement flow
To simplify the reimbursement process for employees, we designed a conversational flow that lives within a DM/chat interface. This flow covers three main intents: Apply, Cancel, and Check Status of reimbursements. The goal was to keep things user-friendly, reduce back-and-forth, and integrate seamlessly with Workday.
1. Apply for Reimbursement
Users initiate the flow by asking for help with reimbursements.
The assistant detects the intent and guides them to:
Upload receipts (with OCR scanning in the background),
Fill out any missing details,
Submit a pre-filled expense form.
A confirmation is shown, and approvers are pinged on slack via Atom (assistant)
Pre-filled forms and receipt scanning reduce friction.
Conversational guidance makes a complex process feel easy.
2. Cancel Reimbursement
When users want to cancel a claim, the assistant lists their expense reports and provides a CTA to go to Workday.
This helps users act fast without digging around for options.
3. Check Reimbursement Status
The assistant checks for submitted reports and tailors the response:
No report found → Prompts to create one.
Single report found → Displays current status (e.g. in progress).
Multiple reports → Lists reports with a select button to view individual statuses.
Flow 2: Request Time Off
We built a chat-first flow for time off requests that:
Detects user intent via natural language.
Connects with Workday to fetch balances and send approvals.
Supports request, cancellation, and status check — all within the chat.
1. Apply for time off
User types something like “I want to take a day off next Friday”.
→ Assistant detects the intent, classifies it, and shows a short form inline in Slack.
→ User picks dates, type of leave (if multiple options exist), and submits.
→ Request is sent to the manager on Workday for approval.
→ User receives a Slack notification once the manager takes action.
Fallbacks in case leave type is unclear or balance is insufficient.
Redirects to Workday only when necessary (e.g., for unsupported cases).
Edge: If multiple dates or types are mentioned, we nudge the user to confirm one.
2. Cancel time off request
User says “Cancel my leave for Friday”.
→ Assistant classifies intent as cancel, shows the current active leave request(s).
→ User confirms cancellation.
→ Request goes to Workday and notifies the manager if needed.
What if no upcoming leave is found? → Assistant gracefully informs the user.
3. Check the status of time off request
We explored auto-listening to status changes from Workday, but this would require constant syncing and deeper infra.
So we decided:
📌 Status is pulled only when the user asks, keeping the setup lean for now.
By bringing time off into Slack, we reduced context-switching, made it faster to take action, and set the stage for more HR flows to live in chat.
SIGNING OFF!
KEY TAKEAWAYS
This project was jam-packed with a lot of work, exploration and discussion with stakeholders. It feels almost impossible to boil down my learning, but if I had to,
here they are!
WHAT THIS PROJECT TAUGHT ME
This project helped me think beyond screens—focusing on clear, conversational flows for Slack. I learned to balance user needs with backend constraints, like syncing leave status only on request to keep things simple and feasible.
Learnings That’ll Stick
Designing for Slack made me more intentional with language and edge cases. I learned how to simplify flows without oversimplifying the user’s needs, and how to collaborate closely with engineers to land on what's doable now vs. what can come later.
Made with <3 and caffeine xD