AI Consulting · Case study
AI-Assisted Product Development · Notion · AnalyticsFrom workflow problem to tested digital product: an AI-assisted Notion build.
How I combined product strategy, AI-assisted execution, analytics architecture and quality assurance to create a customer-ready system.
Submitted to Notion Marketplace October 16, 2026 · awaiting reviewThe challenge
Job seekers often manage applications, interviews, follow-ups and career documents across disconnected tools. A basic tracker can show where an application stands, but it doesn’t necessarily reveal whether the search is generating employer responses, interviews or offers. It can also lose important history when applications close.
I wanted to create something more useful than a pretty dashboard: a free, duplicable Notion system people could actually use to organize their search and understand their progress.
My approach
I led the product strategy and requirements, structured the user experience, and coordinated an AI-assisted development and review workflow.
A key design decision was to measure outcomes correctly. Saved opportunities are excluded from rate calculations. Multiple interviews for one job count as one interviewed application, and canceled interviews don’t count. Closed Outcome fields preserve relevant interview and offer history even after an application leaves the active pipeline.
Where AI added value, and where I led
What I owned
- Product requirements and user experience
- Acceptance criteria for every formula
- Challenging unsupported assumptions
- Human review of what would ship
Where AI assisted
- Iterative implementation
- Formula review
- Documentation and setup copy
- Scenario testing
This was an AI-assisted, human-directed delivery process. It was not an autonomous AI agent, and the free template doesn’t contain AI features.
The validation
Before submission, I tested 16 distinct scenarios across all three analytics formulas. The final test report marked all 16 as passing.
- Employer replies
- Offers
- Withdrawals
- Canceled interviews
- Multiple linked events
- Closed opportunities
- Incomplete records
In a fictional four-application sample, the displayed response, interview and offer rates correctly showed 75%, 50% and 25%. Those are test-fixture calculations, not real job-search performance. I also reviewed the final customer copy for independent databases and internal relations, tested it on iPhone, and duplicated it into another workspace.
The outcome
The completed Job Search Command Center: Free + Analytics includes an application pipeline, connected calendar, analytics, weekly activity and beginner-friendly setup instructions. It was submitted to Notion Marketplace on October 16, 2026 under the listing title Job Search Tracker & Analytics and is awaiting review.
The result at this stage is a tested, reusable digital product taken from concept through submission. Marketplace approval, downloads, adoption and revenue are future metrics, not results claimed here.
Why it matters for my consulting work
This project reflects how I approach AI consulting: start with the business or user problem, design a system around the workflow, use AI where it speeds up execution, and verify the output with measurable acceptance criteria. The objective isn’t simply to use AI. It’s to build something dependable, useful and ready for real people.
Have a workflow that needs structure, automation or better reporting?
Tell me what’s taking too much time or what you can’t see clearly yet. Let’s design a system that works.
Discuss your workflow →