Leading Oracle’s Optical Revolution
As Senior Network Development Manager & Service Owner for OCI Optical Networking, I was responsible for delivering secure and scalable Wave Division Multiplexing (WDM) services across all generations of Oracle’s hyperscale cloud — Gen1, Gen2, and Gen3. With an annual capital budget exceeding $70 million, my mandate was clear: scale faster, operate leaner, and never compromise reliability.
The Career Thread: Optics
This wasn’t my first time building optical infrastructure. The thread runs back to the Bay Area, where I had a hand in building the first DWDM optical ring around the Bay Area with AT&T — early dense wavelength division multiplexing that was bleeding edge at the time. That optical networking experience carried through the Apple iCloud build (carrier-grade routing and optical interconnects) and culminated here at Oracle, where I designed the global backbone.
The Challenge
Oracle’s explosive cloud growth demanded optical infrastructure that could:
- Scale to support 300% year-over-year growth
- Maintain 99.999% availability for nation-state customers
- Deploy new capacity in weeks, not months
- Operate efficiently despite global supply chain disruptions
- Support Gen1, Gen2, and Gen3 architectures simultaneously
Traditional optical network management required armies of engineers for provisioning, monitoring, and maintenance. We had to find a better way.
Oracle’s First Transcontinental Optical Backbone
I designed Oracle’s first transcontinental optical backbone — the long-haul fiber infrastructure connecting Oracle’s data centers coast to coast. This was a significant milestone: Oracle building and owning its own optical backbone rather than leasing capacity from carriers. Led the global optical plant refresh alongside this, bringing the entire worldwide optical infrastructure to a consistent, automated standard.
The Automation Breakthrough
844 Hours Saved Annually
I led the development of an optical automation platform that transformed how we operated:
Automated Provisioning
- Zero-touch wavelength deployment
- Automated power optimization
- Self-configuring amplifier chains
- Intelligent route computation
Predictive Maintenance
- Optical signal degradation prediction
- Proactive component replacement
- Automated failover orchestration
- Self-healing fiber paths
Capacity Planning Intelligence
- ML-driven demand forecasting
- Automated inventory management
- Supply chain optimization
- Just-in-time deployment strategies
$27 Million Revenue Acceleration
The real magic happened when we applied this automation to datacenter deployments:
- Before: 16-week optical deployment timeline
- After: 6-week fully automated deployment
- Impact: 10 weeks faster time-to-revenue per datacenter
This acceleration directly resulted in $27 million in additional revenue by bringing capacity online faster than competitors.
Technical Innovation
The Stack
- Control Plane: Custom Python automation framework
- Monitoring: Optical performance monitoring (OPM) integration
- Analytics: Time-series analysis for predictive maintenance
- Orchestration: Ansible for configuration management
- APIs: RESTful interfaces for cross-platform integration
Manufactured Rack SKUs
Developed 4 manufactured integrated rack SKUs — standardized, pre-built optical equipment racks that could be deployed consistently across Oracle’s global data center footprint. Combined with a white box switching platform achieving 100Gbps, these racks turned optical deployment from a bespoke engineering project into a repeatable manufacturing process.
Key Innovations
- Optical Digital Twin: Created virtual models of the entire optical network for testing changes before deployment
- API-First Design: Every optical element exposed via APIs for programmability
- GitOps for Optics: Configuration as code for all optical systems
- Closed-Loop Automation: Self-optimizing optical paths based on real-time performance
Navigating Global Challenges
During my tenure, we faced unprecedented supply chain disruptions. My team’s response:
- Re-engineered deployment processes for component flexibility
- Created vendor-agnostic automation frameworks
- Built strategic inventory reserves using predictive analytics
- Established alternative supply chains across three continents
Result: Zero deployment delays despite global chip shortages.
Impact Beyond the Numbers
For Oracle:
- Established optical networking as a competitive advantage
- Created IP that influenced Oracle’s entire infrastructure approach
- Built a team that became the gold standard for infrastructure automation
For the Industry:
- Proved that optical networks could be as programmable as servers
- Set new benchmarks for capital efficiency in optical deployments
- Influenced vendor roadmaps toward automation-friendly designs
Leading Through Transformation
This role taught me that infrastructure leadership isn’t just about technology—it’s about:
- Building teams that can innovate under pressure
- Turning constraints (budget, supply chain) into innovation catalysts
- Creating systems that scale beyond your tenure
- Balancing fiscal responsibility with technical excellence
The principles I developed here—automation-first thinking, API-driven infrastructure, and lean operations—continue to influence how I approach every technical challenge, from AI systems to supply chain optimization.
The Bottom Line
With a $70M budget and a small team, we didn’t just build optical infrastructure—we revolutionized how Oracle thinks about infrastructure automation. The 844 hours we saved annually was just the beginning. The real victory was proving that with the right approach, infrastructure can be a profit center, not just a cost center.