CFO
Needed a real-time view of close health and forecast risk so they could quickly decide where attention was needed during period close.
Oracle NetSuite
Designing a faster way for finance teams to review month-end close
I designed and shipped a new Oracle NetSuite review experience for month-end close. It replaced a static checklist with a clearer workflow that helps finance leaders spot blockers, assess risk, and act faster.
Challenge
Month-end close review was spread across multiple tools.
Shipped
A 0-to-1 Close Manager workflow inside Oracle NetSuite.
Changed
One guided review surface for blockers, risk, and task assignment.
Role
Owned the core UX from concept exploration through handoff.
Context
Month-end close is a high-stakes workflow with a short decision window.
Finance leaders need fast, reliable answers about what is blocked, what is material, and where to act next.
Timeline
3 months
Team
PM, engineers, and finance SMEs
Users
CFOs, controllers, and accountants
My role
I owned the core UX from concept exploration through developer-ready handoff. I drove the shift from a static checklist to a real-time review experience, aligned the direction with PM, engineering, and SMEs, and delivered the final interaction model and specs.
Scope
Review experience, AI summaries, workflow states, handoff specs
Controller
Needed to manage and expedite period-close tasks, identify exceptions quickly, and keep financial reporting accurate and on schedule.
Accountant
Needed clearer task ownership, reminders, and due-state visibility so work could be completed on time without relying on manual follow-up.
Status quo
Before Close Manager, the workflow buried the next action inside manual coordination.
Finance leaders were checking status in multiple places, chasing updates, and piecing together what was holding the close up instead of moving it forward.
Before Close Manager
Finance leaders were stitching close status together manually before they could decide what to do next.
Step 1Manual checklistClose review started in a static checklist instead of a live operational view.
Step 2SpreadsheetsStatus and variance context lived across separate files and reports.
Step 3Status chasingFinance leads followed up manually to learn what was actually blocked.
Step 4Delayed actionImportant issues surfaced late, when there was less time to respond.
Status was spread across spreadsheets, manual reviews, and category systems.
No single view showed close health and next action together.
Manual coordination increased as categories and exceptions grew.
Process
I kept the public process to the three design moves that most changed the outcome: separating insight from action, improving comparison, and then resolving scale.
Initial direction
The first concept separated insight from action.
The initial direction looked comprehensive, but it made users connect risk to action on their own.
Overview, insight, and task detail all lived in separate zones, so finance leaders had to scan back and forth to understand what actually needed action.
Overview
Close health was visible, but too detached from the work that needed attention.
Insight
The summary sat beside the flow instead of guiding the next step.
Action
Users still had to translate risk into tasks on their own.
Explorations
The next concept improved overview, but only for two categories.
The two-column version made payables and receivables easier to compare side by side, and it brought summary and detail closer together. But it depended on the screen staying small and symmetrical, so it could not scale once more categories and richer AI states were added.
What it improved
This was the first direction that made progress easier to compare, but it still did not establish a scalable review flow.
Scalability issue
As categories grew, the review experience became harder to scan.
Adding more category cards and parallel AI summaries increased visibility, but it flattened hierarchy and made the next action less obvious.
Clutter
Too many categories competed at once.
Parallel summaries
AI insight no longer had a clear place in the flow.
Constraint
More visibility was useful, but not if the next action disappeared.
Solution
I designed one place for finance leaders to monitor financial health, see real-time priorities, and assign work.
The experience brought close health, the highest-priority issues, and task assignment into one workflow so teams could close the books on time without the usual end-of-month stress.
Global status
Overall close health stayed visible.
Selected category
The middle column kept the current workstream in focus.
Actionable detail
Task detail stayed adjacent to the summary.
Key product decision
Use AI to prioritize and explain risk, not automate financial decisions.
This gave finance leaders speed without losing trust or auditability.AI-generated insights
AI summaries helped surface what needed attention first.
The AI did not make the decision. It helped finance leaders understand where to look before they took action.
Trust model
The AI pointed to what mattered first. Finance leaders still verified the source and made the decision.
AI summarizesHighlights likely blockers, risk signals, and the category needing attention.
User verifies sourceFinance leaders can inspect linked reports before moving the workflow forward.
User actsHumans keep control of task assignment, escalation, and financial decisions.
Financial report data from accounts payable and receivable, along with close-status signals from category workflows.
A concise summary of close health and the specific areas that needed attention.
Finance leaders could verify the summary through linked report sources before taking action.
Because these were high-stakes financial decisions, the AI highlighted what needed attention while the decision and action still stayed with finance leaders.
Impact
A shipped 0-to-1 workflow replaced the old static close review.
The outcome was a live review experience that gave finance leaders one place to assess close health, understand risk, and move work forward before the end-of-month scramble.
Close Manager workflow shipped inside Oracle NetSuite
for close health, priorities, and task assignment
AI guidance stayed explainable before finance leaders acted
Proof 01
Replaced a manual multi-place checklist with a live review workflow.
Proof 02
Made blockers, ownership, and next action visible in one workflow.
Proof 03
Established a scalable pattern for AI guidance in an auditable enterprise product.
Learnings
What this project taught me
This project taught me that the hardest part of enterprise AI design is not adding more information. It is deciding what should stay visible, what should collapse, and what needs to become action.