Before
- 01Prepare operator data
- 02Assign routes manually
- 03Reconcile changes
- 04Complete operational setup
01 · B2B SaaS · AI-assisted operations
Making automation useful—and trustworthy—for dispatchers.
I owned the product vision, roadmap, and MVP for a cross-platform scheduling SaaS. The central product challenge was not adding AI for its own sake; it was reducing manual route work while preserving operator control.
85%of manual route assignments automated
Days → minutesoperator onboarding after bulk import
Ahead of planMVP delivery
The situation
Manual assignment was the bottleneck, but blind automation was not the answer.
Dispatch workflows involved repeated manual route assignments and operational setup that could stretch across days. Feedback from dispatchers became product requirements and exposed the workflow bottlenecks the MVP needed to remove.
My responsibility covered the whole product loop: shaping the vision, sequencing the roadmap, deciding what the first useful release needed, and defining the events required to measure adoption and friction after launch.
Shipped evidence
The product is live on the web and published on the App Store.
Workflow decision
Move repetitive work to the system. Keep consequential judgment with the operator.
Product direction
- 01Bulk import
- 02AI-assisted scheduling
- 03Validation gates
- 04Operator review and billing
Three consequential decisions
The roadmap was a sequence of risk reductions.
Put validation around automation
The scheduling workflow used validation gates so automation could remove repetitive assignment work without asking operators to accept an unchecked result.
Prioritize bulk import
Billing and bulk-import experiences reduced setup from days to minutes, shortening the distance between interest and a usable operational account.
Make build-versus-buy explicit
I evaluated multi-tenancy and API decisions against scale, delivery speed, and compliance instead of treating architecture as separate from product strategy.
Measurement
Define the learning loop before calling the product finished.
I defined product analytics and event tracking to measure adoption, expose friction, and guide iteration. The public evidence available today is strongest around workflow automation and onboarding time; retention and funnel data are not published.
Public case study note
Product screens and public release links are shown above. Internal analytics, customer identities, and confidential operational data are intentionally omitted.
What this changed in my product practice
“Automation earns adoption when the product makes control, exceptions, and confidence part of the workflow—not an afterthought.”
If I were extending the case study now, I would pair the automation metric with validation-failure rates, time-to-first-schedule, and cohort retention to show not only that work moved faster, but that the new behaviour persisted.