Case studies
Work that held up under real-world complexity
A few examples of how I approach messy, high-stakes systems — where the work is not just shipping features, but creating structure that operators, customers, and technical teams can actually use.
Scaling device lifecycle systems for 9M+ subscribers
Built and scaled a platform ecosystem supporting 9M+ subscribers and $100M/month in device inventory flow — reducing support load by ~90% through better system design, automation, and operational tooling.
The situation
Device operations were fragmented across systems, teams, and workflows. Support teams were often compensating for gaps in the platform — using manual workarounds, tribal knowledge, and disconnected tools to keep the business moving.
At small scale, that kind of operational friction is annoying. At millions of subscribers and high-volume device flow, it becomes expensive, risky, and difficult to manage consistently.
The real problem
The issue was not simply that teams needed more features. The deeper problem was that the systems were not behaving like one coherent operating model.
Every disconnected workflow created another place for support cases, manual intervention, inventory risk, customer friction, or operational ambiguity to leak into the business.
What I changed
- Unified fragmented device lifecycle workflows into a more coherent platform layer.
- Introduced automation across provisioning, returns, swaps, and support-heavy operational paths.
- Integrated external systems like Apple GSX and eSIM provisioning into internal operational workflows.
- Designed tools that gave operators clearer visibility, better context, and fewer reasons to escalate.
- Translated business, technical, and operational constraints into product decisions that could scale.
The outcome
The platform supported device lifecycle operations across 9M+ subscribers and $100M/month in inventory throughput, while reducing support cases by approximately 90% through automation and better system design.
Just as importantly, the work shifted the operating model from reactive support to structured workflows — giving internal teams a more reliable way to manage high-volume device events.
Why it worked
The key was not adding complexity to the system. It was removing friction between the business process, the technical architecture, and the people operating it every day.
This case reflects how I typically operate: find the real system failure, translate across teams, and build something that survives contact with reality.
See how I operate →Additional proof points
A few smaller snapshots that show the same pattern across fraud detection, internal tooling, and telematics: find the signal, structure the system, and make the work operational.
Finding fraud patterns hidden in device behavior
Account takeover fraud was happening at scale with no systematic detection in place. The signal was there — buried in device behavior patterns — but it had not been translated into an operational detection method.
- Identified behavioral clusters tied to account takeover patterns
- Built detection logic using device activity signals
- Surfaced linked accounts to enable proactive intervention
- Worked with fraud and ops teams to operationalize findings
- $10K+/wk in fraud prevented
- 140+ linked takeovers identified
- Moved the team from reactive review toward proactive detection
SCOUT — internal operations platform
Internal operators across supply chain, customer care, and device operations were working through disconnected tools with limited shared context. SCOUT created a cleaner operating interface for high-volume support and lifecycle workflows.
- Led product definition and roadmap for the internal ops platform
- Consolidated fragmented workflows into a unified operator interface
- Built automation layers to reduce manual intervention points
- Designed around the real behavior of 2,500+ internal users
- 2,500+ operators on platform
- Reduced context-switching across support workflows
- Embedded process knowledge directly into the product experience
Turning telematics data into fleet behavior change
Motovate needed to turn raw telematics data into something useful for drivers, fleet managers, and enterprise buyers. The challenge was translating signals into behavior insights that were accurate enough to trust, clear enough to act on, and valuable enough to sell.
- Defined product logic for translating telematics signals into driver scores
- Built feedback loops between behavior data and coaching interventions
- Designed fleet manager dashboards for actionable oversight
- Sold and positioned the product with real fleet operator accounts
- Improved driver behavior scores over time
- Adopted across multiple fleet operator accounts
- Built a practical foundation for future fleet safety and behavior-change contracts