Work / Manufacturing Scheduling Application
Paused
Manufacturing Scheduling Application
One shared schedule to replace three spreadsheets, designed and tested with the schedulers, sales reps and operations leaders who'd use it.
Problem
The factory's schedule lived in three spreadsheets. Sales called the scheduler before promising dates, and operations heard about delays too late.
My Role
Led UX and prototyping on a team of 2 UX designers, 3 business analysts, and 1 project manager · 8 weeks, 2026
Tools
Survey · Figma · Figma Make (Figma's AI tool that builds clickable prototypes) · In-person interviews
Outcome
Research and a clickable prototype, shaped over 60+ versions with the people who'd use it. My way of prototyping with AI was adopted for 3 more projects. Paused before it was built, when company priorities changed.
The short version
Before: The schedule lived in three spreadsheets. Sales had to call the scheduler before promising a date, and conflicts often showed up after the customer was told.
After (designed, not built): One shared schedule that still works like Excel. Sales and operations can check it themselves, and people get an alert when a date they're tracking changes.
Why read a paused project? Company priorities changed before it was built, so there are no launch numbers here. What it does show is one of the hardest parts of any new tool: getting people who trust their spreadsheet to want to switch. And the team adopted my way of prototyping for 3 more projects.
From "cold, dead hands" to "I can see myself using this"
At the start of one interview, the master scheduler joked that we'd have to "pry Excel out of my cold, dead hands." So I changed the prototype live as we talked, keeping what she relied on in Excel: moving projects freely and trying out changes before everyone else sees them. By the end, she said she could see herself using it.
What people told us
I started with the spreadsheets and documents people gave us, plus a seven-question survey of 30 employees who know the scheduling process well.
From those, I made wireframes, and Figma Make turned them into a first clickable prototype. Then the business analysts and I ran three days of in-person group interviews, changing the prototype live as people reacted.
Three things stood out:
People wanted to keep what they knew. Excel supported habits people relied on and didn't want to lose.
Changes didn't reach people in time. Schedule changes and project status didn't always reach the people who needed them soon enough.
One view wouldn't fit everyone. Schedulers, sales and operations each used the schedule differently.
Who uses the schedule
I wrote up three profiles, one for each role the spreadsheets hurt most: the scheduler who keeps the schedule current, the sales rep who promises dates from it, and the operations leader who watches how busy the plant is. They guided what each person sees in the design.
Master scheduler
Keeps every project on track across several building lines, and needs to change the schedule fast without breaking it.
Needs to
Work from one current version of the schedule
Move projects without retyping the sheet
Compare how much each building line can take on
Pain points today
The schedule lives in three different spreadsheets
Retypes rows every time a date shifts
Chases people to confirm they saw a change
What the design gives them
One central schedule that's always current
Drag-and-drop editing that works like the spreadsheet habits they already have
Sales representative
Quotes delivery dates to commercial clients, and needs to trust that date before promising it.
Needs to
Check what the plant can really build before quoting a date
Know right away when a tracked order shifts
Get answers without interrupting the scheduler
Pain points today
Has to call the scheduler for basic status checks
Learns about conflicts after the customer was told
No simple view built for a sales role
What the design gives them
A view-only copy of the live schedule, anytime
Automatic alerts when a tracked order changes
Operations leader
Responsible for on-time delivery and how much the plant takes on, and needs to spot risk early without digging through every row.
Needs to
See plant-wide status at a glance
Catch at-risk projects before they slip
Compare how busy each building line is
Pain points today
Asks the scheduler instead of checking directly
Finds out about delays after they happen
Stitches separate sheets together to compare lines
What the design gives them
A dashboard with the key numbers for active projects and plant workload
Early flags on overdue tasks, before they become delays
Each profile blends several real people from the survey and interviews with schedulers, sales and operations.
Design
The first prototype brought the schedule into one app, without losing what people liked about Excel. It focused on:
One place for the schedule and project details
Spreadsheet-like editing, plus drag and drop
Changes and project status that are easy to spot
A first dashboard with the key information
Testing and changing it
I gave Figma Make my wireframes and a detailed written brief, including our team's design system, and it built the clickable prototype. That let me make new versions fast.
Each week I put the latest version in front of teammates and the people who use the schedule, then changed it based on what they said. Over 8 weeks, that added up to more than 60 versions.
Where the AI's first guess went wrong
My wireframes came from the team's own spreadsheets and documents. Where they left gaps, Figma Make filled them with general ideas about how scheduling works. But this team's process was customized in ways general ideas don't cover. One example: it added cards of extra numbers under the drag-and-drop scheduling screen. They repeated numbers the screen already showed, and the schedulers didn't care about them. They just wanted to get their job done. Leadership wanted those numbers, so I moved them off the scheduling screen and into a metrics view.
I didn't treat the AI's version as the answer. I used its general ideas, plus our research, to start the conversation in the interviews. There, people confirmed what worked and showed us what needed to change.
The choices we made
We considered: Staying in the spreadsheets
We chose: A new, easier scheduling screen that keeps the Excel habits people relied on: moving projects freely and trying out changes before everyone else sees them
Why: People trusted their spreadsheet. The new tool had to feel familiar to be worth switching to.
We considered: One fixed dashboard for everyone
We chose: A dashboard each person sets up
Why: Each role needed different information.
Making the AI budget go further
Figma Make limits how much AI work you can do, counted in credits, and I hit that limit quickly. So I wrote tighter prompts, and we put AI prototyping first on the screens that would gain the most from it. I also made the case for more funding for AI prototyping, and got it.
How we knew it was getting better
In the interviews, we asked people where they would go to do a task. They pointed to the spot on the screen and told us what they'd pick. As the versions went on, they found what they needed easily, and they told us it looked like a tool they would use. One example: early on, a project's system messages sat on a separate Communications page. By the later versions, people could find everything about a project in one place: its details, plus all its related documents and conversations.
Figma Make's first try, made from my wireframes, a written brief and our design system.
Screens are studio redraws. Names and data are invented.
I made it easier to move between the dashboard and the Projects page. Then I let each person choose and arrange what their dashboard shows, because different roles needed different things.
In interviews, people said system messages would be more useful inside each project than on a separate Communications page, so I moved them there.
The new scheduler: drag a project onto the schedule instead of typing it into Excel.
Design Outcomes
Adopted for 3 more projects
The Senior Manager of Digital Products made my way of prototyping with AI part of the team's process and used it on 3 larger projects.
"I can see myself using this"
The master scheduler, by the end of one interview.
60+ versions in 8 weeks
Weekly testing with the people who use the schedule turned one fixed dashboard into one each role can set up.
"Her willingness to learn new tools and techniques, including AI-assisted prototyping, has helped our team improve collaboration with stakeholders and accelerate the design process."
Melanie Morgan · Senior Manager of Digital Products
Learnings
Modernizing doesn't mean starting over. Keep the habits that work, even when the tool underneath changes.
Flexibility was a must, not a nice-to-have. Each role needed different information, so one fixed view wouldn't do.
People find needs when they can react to something real. Changing the prototype live in interviews let us check what people needed on the spot.
If the project resumed
Test drag and drop and the set-it-up-yourself dashboard with people doing real tasks on their own: are they faster than Excel?
Pick the smallest first version worth building, based on what helps people most, what's doable to build, and what the business needs.
Turn the design into reusable building blocks and clear rules, so developers build it the same way every time.
Decide how to measure success: time spent scheduling, manual steps, errors, how many people use it, and how fast people see changes.
Have a tool that needs this kind of attention? Let's talk about it.