Skip to content
HeyKaylz.

Case study 01 · Workflow Digitization

From email chains to a self-serve approvals portal

Operations staff were processing shared task updates through email threads and shared spreadsheets. I helped turn a fragile manual process into a tracked, self-serve workflow — starting with the people doing the work.

Client
Fortune 500 client · Public Service Industry
My role
Technical consultant · requirements & stakeholder lead
Timeline
3 Months
Team
2 engineers, 1 designer, client operations leads

40%

faster request cycle time

100+

manual handoffs removed

33%

fewer rework errors

01

The problem

Task handover passed through 3 various channels depending on who was sending it. Requests lived in inboxes, status lived in a spreadsheet, and nobody could answer the simplest question: “Where is my request?”

The business cost was real — average turnaround of 30 days for simple requests, frequent re-submissions due to missing information, and no audit trail for compliance reviews.

02

Discovery

Before a line of code was written, I shadowed the operations team and mapped the journey end-to-end. Three insights reshaped the project:

  1. Most delays weren't approvals — they were missing information. Roughly 60% of requests bounced back at least once.
  2. Approvers didn't need a new tool; they needed fewer interruptions. A daily digest mattered more than a dashboard.
  3. Compliance was the silent stakeholder. An audit log was a must-have, not a nice-to-have.

03

Scoping the MVP

The original ask was a full workflow engine. I proposed a narrower first release focused on the highest-volume request type, which let us validate an MVP approach in 6 weeks, beating the 3 month complete timeline.

In the MVPDeferred to v2
Smart intake form with validationConfigurable multi-branch routing
Rules-based routing for one request typeMobile approvals
Approver queue + daily email digestAnalytics dashboard
Immutable audit logSelf-serve form builder

04

The solution

  1. 01Intake formRequired fields enforced up front
  2. 02Validation rulesCatch missing data before routing
  3. 03Routing engineAssigns approver by type & threshold
  4. 04Approver queueDaily digest + one-click decisions
  5. 05Audit & syncLogged and pushed to ServiceNow with MuleSoft
High-level request flow

I wrote the user stories and acceptance criteria, partnered with our designer on the intake flow, and managed the integration component with ServiceNow so engineering could build in parallel.

05

Trade-offs & alignment

Leadership wanted every request type on day one. Instead of arguing, I brought data: the top request type represented 70% of volume. Shipping it first meant real differnce within weeks — and a pattern we could copy for the rest.

06

Outcome

  • Average cycle time dropped from 30 days to 10 days.
  • Re-submissions fell by 33% thanks to up-front validation.
  • Compliance reviews went from a week of spreadsheet archaeology to a 2 minute export.
  • The pattern was extended to 2 additional request types in v2.

07

What I'd do differently

I'd instrument the old process before replacing it. We had to reconstruct our “before” metrics from email timestamps — a baseline captured on day one would have made the impact story faster to tell and harder to dispute.