PORTFOLIO
Resume
Executive Summary
Data and Observability Specialist with 20+ years engineering data systems and infrastructure that hold up at scale. I speak both Engineer and Executive – fluently.
I operate in three core areas: I untangle messy business and technical problems into clear requirements, I engineer solutions which stand the test of time, and I translate between engineers and executives.
Results include: identified $100,000/mo in ghost services, dropped SaaS bill by $75,000/mo, cut Intranet downtime by 99% at a Fortune 500, and scaled a metric pipeline past 13M samples/sec.
My daily work is large-scale databases and high-performance pipelines under strict business rules. Elegantly solving hard problems is my happy place.
Work History
(More Details in Project Summaries Below)
Sr. Site Reliability Engineer – Data & Observability
Kaseya
2019 – Present
Engineered and supported an internal observability service, reducing cloud cost by 80%. Wrote ETLs, built dashboards, and engineered decision supports systems, culminating in over $500,000/mo in cost savings from direct contributions.
Sr. Infrastructure Architect – Automation & Analyst
O’Reilly Automotive
2011 – 2018
Automated, standardized, and scaled Intranet and Online Store systems, reducing downtime by 99% on systems operating at $50,000/minute of sales. Dedicated 50% of time to advising and mentoring.
System Administrator – Jack of all Trades (SMB)
Cart32
2005 – 2011
Rebuilt systems and networking to meet PCI compliance and reduce downtime. Managed audits for 3 years. Restructured data center and contracts, reducing costs by 40%.
(See more in Project Summaries below.)
Education Background
Master’s Degree – Data Management
Dakota State University
Madison, SD
2016 – 2017
Bachelor’s Degree – Info Systems
Missouri State University
Springfield, MO
2003 – 2007
Stack Dump
Infrastructure & Environment
Remote. Physical and virtual systems. On-prem, semi-cloud. VMWare, OpenStack, and AWS. Linux, some Windows.
Data & Observability
PostgreSQL and MySQL RDBMS. VictoriaMetrics & Prometheus TSDB. Grafana. Telegraf. SQL, PL/SQL, promQL.
Programming & IaC
Python, C++, Ruby, Bash, and PowerShell. Puppet, Ansible, and Terraform. HTML, CSS, and REST APIs.
Personal Sidebar
I Love To Dive
I Love To Drive
Project Summaries
VMWare Decision Support System
Stakeholder Request
Summarize global assets and performance across 19 vCenters for executive review and cost optimization (CoGS analysis).
Solution
A Powershell ETL (client’s language preference) to extract VMWare objects and metrics. Data stored in Postgresql. Dashboards added to customer’s Grafana. Automated, daily update cadence.
Outcome
A single pane of glass across a global fleet spanning 14,000 assets and 120 petabytes of storage. Massive drawdown on a $5M/mo savings goal. Better price alignment of services.
Executive Advisor and Mentor
Stakeholder Request
Advise executives on large-scale, fleet-wide technology decisions. Allocate 50% of my FTE time to mentoring, advising, and training Junior, Mid, and Senior SREs.
Solution
Deep dives into technology requests from leadership. Recommendations pushed upward with evidence-backed reasoning. Held weekly training sessions for SREs and developers on existing solutions, technologies, and future directions.
Outcome
A dedicated translator between technical and non-technical (executive) audiences. Improved clarity and alignment between business and technical goals (and realities!).
Metric Pipeline at Millions/sec Scale
Stakeholder Request
Scalable pipeline to consume over 1,000,000 messages (13M samples) per second, replacing unsupported legacy system.
Solution
Engineered an Ansible playbook (client’s existing IaC) to install Telegraf agents in lieu of monolithic, legacy software. Tuned Kafka partitions and workers. Low cost. Zero licensing (open source). Stateless. Infinitely scalable.
Outcome
Reduced hardware requirements by 20% compared to legacy system. Eliminated 100% of config drift.
Cluster-Aware Patching System
Stakeholder Request
Bespoke, cluster-aware patching of 450+ systems. Automated.
Solution
Wrote a semi-modular framework in Bash and Python to align to their stack. Version controlled. Dedicated logging. Metric output. Alerting possible (external), but not in-scope. Fully automated.
Outcome
Reduced human workload by 80%, dropping from ~4 FTEs to 1. Eliminated 95% of patching errors and related downtime.
Salesforce Revenue Leak – Ghost Services
Stakeholder Request
Untangle and clarify Salesforce billing data (contracts), and identify unpaid assets and services for recovery.
Solution
Fleet-wide scanner to find, interrogate, and catalog all assets into a Postgresql database. ETL to codify muddy, flimsy sales rules in Salesforce into logical correlation against assets in the same PG DB. Grafana dashboards for executive and operational consumers.
Outcome
Revoked 8 petabytes of storage from non-paying accounts, valued at $100,000/mo. Established importance of data hygiene, leading to tighter controls in Salesforce.
CDN Cost Explosion
Stakeholder Request
Find and optimize Akamai (CDN) spend on flagship product after explosive rise in usage and cost over 18 months.
Solution
Manual analysis of endpoints, focused on their Pareto distribution (80/20 rule). Found and submitted a patch for a Web Request in their application’s updater logic to use proper HTTP headers to detect changes; not blindly download files every hour. Drafted proposal for CloudFront alternative.
Outcome
Reduced CDN bandwidth by 95%, valued at $5,000/mo. Reduced vendor sprawl by one, replacing Akamai for CloudFront (client already had AWS assets).
DataDog Cost Analysis Engine
Stakeholder Request
Design meaningful reports on DataDog spend for attribution to products, teams, and applications.
Solution
A Python ETL (team language preference) to read the usage API and extrapolate spend, allowing faceting by tags. Grafana dashboard highlighted spend by-tag (product, team/owner, application). Executed daily.
Outcome
Reports drove accountability up and highlighted $75,000 of monthly savings. Client executed slowly, but used historical metrics to draw down on costs and drive new policy.
Application Deployment Pipelines at Scale
Stakeholder Request
Wrangle 80 Java applications behind company Intranet to address growing pains, instability, and inefficiency.
Solution
Rebuilt Tomcat and Apache HTTPD systems to be cluster-aware, fault tolerant, and horizontally scalable. Wrote a deployment program (Bash) to synchronize, configure, and standardize applications across the fleet. Automated everything with Puppet (client’s IaC).
Outcome
Took downtime from 1-2 hours/day to virtually zero. Reduced config drift by 100%. Reduced deployment time by 50%. Better config hygiene reduced data source contention. Application isolation aided problem identification and drastically reduced MTTD/MTTR. Client pushed application count over 200 and web volume up by 300% without issues.