Operations leadership with current cloud depth.

Situation

Customers ran the company's association-management software on their own infrastructure. Every install was different, support was expensive, and cloud-native competitors were starting to pull customers away. Hosting also meant a higher bill for the same application, so the move had to make sense to every audience.

What I did

• Made the case for a hosted, private-cloud delivery model and led its rollout with a distributed team of up to 12
• Turned one-off migrations into a repeatable playbook the team could run
• Showed each audience its value: lower total cost for executives, remote access for staff, API access and dashboards for data, a refreshed portal for members
• Led the API initiative, with read-only access to 1,500+ data fields, adopted by about 70% of customers
• Reported progress and risk to executive leadership

Result

• 115+ organizations and 1,300+ users on hosted delivery
• Customers kept software they knew without running servers, which helped keep them from leaving
• Supported hosted revenue growth from about $1.6M to $2.6M over five years

Situation

A cloud-first SaaS platform running dev, QA, beta and production environments in AWS, where spend had grown along with them.

What I did

• Reviewed utilization and rightsized over-provisioned resources
• Planned reserved-instance and commitment purchases against real usage
• Cleaned up unused resources
• Put governance practices in place so the savings held

Result

• Roughly 32% lower AWS spend
• Savings kept through ongoing governance, not a one-time cleanup

Situation

Production SaaS with frequent releases, a steady flow of cloud security findings, and an on-call rotation that needed better signal.

What I did

• Proposed blue/green and canary strategies from 2024 AWS workshops and won approval for blue/green; the team moved releases to production validation with fast rollback, then automated the steps
• Worked Wiz and Ionix findings with security and engineering: prioritize, track, resolve
• Administered CrowdStrike Falcon; worked with AWS WAF and Shield
• Tuned CloudWatch and PagerDuty alerting; automated remediation with SSM and CloudFormation

Result

• Every release had a tested way back
• Zero open critical findings, with new ones resolved within SLA
• Releases no longer tied to late-night windows
• Less manual intervention and faster recovery during incidents

Situation

Backend services degraded over time. The usual response is to scale up and move on: bigger instances, more of them, a higher bill, and the problem comes back anyway.

What I did

• Correlated performance metrics with CloudTrail events and instance logs to line up the timeline
• Traced memory growth patterns back to the backend code behind them
• Handed developers a clear, evidence-backed root cause instead of a vague "it's slow"

Result

• Memory leaks fixed at the source by development
• Avoided throwing more infrastructure at a software problem
• Better working trust between ops and engineering