Case study

Release notes and revenue-sharing automation

Two internal tools I designed at K2U, and how I cut repeated work for the team.

Result80-90 days to 30-45; 1 hour to 15 minutes
My roleProduct lead: concept, design, approval
TeamEngineering, content, external experts
CompanyK2U

Release-notes workflow

The problem

K2U courses teach Salesforce, and Salesforce ships three releases a year. After each release, someone had to read the release notes, work out what changed, and decide which courses needed updating. Update cycles ran 80 to 90 days, which meant learners could be studying outdated material for months.

I ran each release cycle by delegating the release-note reviews and update projects across the content team and project managers, then reviewing the changes and pushing the updates live myself. Reading the notes was the slowest part.

What I designed

An AI workflow that reads the release notes and flags which course content needs updating. I led it from idea to approved build: user interviews, wireframes, flowcharts and the business case to management. The engineering team built it to my design.

1Release notes published
2Workflow reads them
3Flags affected courses
4Team updates the content

Result

80-90 → 30-45

Update cycles fell from 80-90 days to 30-45 days.

Revenue-sharing tool

The problem

We paid external experts a share of their course sales. Each month I spent about an hour per expert building their report, across about 8 experts.

What I designed

I designed the concept and how it would work, using the Metorik API for sales data. Engineering built it. Each expert now sees their own sales and refunds by date, and generates their invoice from our template.

1Sales data from Metorik
2Each expert sees their sales and refunds
3Invoice from our template

Result

1 hour → 15 minutes

Monthly reporting per expert fell from about 1 hour to about 15 minutes, and experts got self-service access to their own numbers.

Onboarding and SOPs

I led a team of 18 (15 in content and 3 project managers) plus about 8 external experts. I redesigned onboarding and wrote the procedures, cutting ramp-up from about 3 weeks to about 1, with experienced teammates running the walkthroughs.

After the acquisition the team shrank to 5. I kept delivery going by automating work instead of asking for replacements.

How I decide what to automate

I look for work that repeats on a schedule, takes the same steps each time, and depends on one person remembering to do it. If all three are true, I write it down first, then decide whether a procedure, a template or a tool is the right fix. The tools above came from that approach, and so did the onboarding procedures.

Skills and tools used
Process designAI workflow designMetorik APISOPs and onboardingStakeholder pitchingTeam leadership