Austin, Texas Metropolitan Area
Founding-engineer instincts. Operator's discipline. Manager's eye for leverage. Twenty years in, the throughline is range. I have built and led infrastructure across domains that punish bad engineering the hardest: institutional trading and low-latency finance, two crypto exchanges, a consumer platform that ran as the fifth-largest S3 customer on Earth, virtual cyberwarfare ranges for the DOD and intelligence community, and now identity verification at SentiLink. Different stacks, same job. Make it reliable. Make it cheap. Make it ship. A few things people consistently get from me. I take cost seriously: I have led initiatives that have removed nine figures of cloud spend over my career, including a single EMR redesign that cut a client's bill from $16.8M to $3.2M. I take uptime seriously: I held a 99.99% SLA at TradeBlock through a live multi-account, multi-region carve-out from a sister company, and passed two SOC 2 audits in the middle of it. I take people seriously: every engineering team I have run, I have hired and grown myself, and I would rather develop a strong engineer into a great one than do the work alone. I am at my best on hard, ambiguous problems at the bottom of the stack, the ones where reliability, cost, and developer leverage are all in tension and the right answer is not obvious. I am most useful when the stakes are real. Specialties: AWS and GCP at scale / Kubernetes / Terraform and Terragrunt / Spacelift / CI/CD / SRE and incident command / Cost engineering / Multi-region and multi-account architecture / Low-latency finance / Crypto and digital asset exchanges / ASIC and AI infrastructure / Team leadership
A cross-chain intent-centric orchestration, settlement, and asset issuance network. The primary focus is cross-chain lending/LVR-free liquidity which launched with lend.finance, Nucleus, and GoldLink
TradeBlock's API-centric end-to-end platform enables institutions to streamline their crypto asset trading workflows with purpose-built pre-trade, trade execution, post-trade, and reporting solutions. * Responsible for multiple engineering teams including SRE, Corp IT, Level 2 Support, and Database Engineering. Inherited and completely revamped an aging and overburdened backend platform. Developed a new automated SDLC deployment system using Clair, Rectory, ArgoCD, and K8. • Redesigned the original infrastructure which included antiquated and severely deprecated platform versions. Generated all infrastructure into code using Terraform and Ansible. Scaled and hardened the original platform by creating multi-AZ and Multi-region failover. • Hired and trained teams of engineers for the Platform, FIX, and MarketData teams. Was involved in the hiring process for every engineer since starting with the company. This includes hiring and training all engineers for the spin-off company CDI. • Successfully complete two separate SOC II Type 1 and 2 audits. The first audit was successfully completed for CDI while in the middle of spinning out Tradeblock trading from the parent company. • Maintained a 99.99% uptime SLA while separating all the services from the sister company during the split. This included migrating live systems between accounts, VPCs, and Regions while maintaining availability. Was fully responsible for the planning and implementation of the split and our subsequent handoff.
Spry is a leading DevOps consultancy specializing in revitalizing and re-engineering DevOps practices for financial services companies. We implement and integrate cloud-native technologies appropriate for company-wide DevOps transformations. * Responsible for all client-facing interactions, code planning, design reviews, scope planning, and implementation. Worked with C-level executives to plan and construct migration plans spanning multi-year projects. Total AWS/GCP spend reduction has crossed 200M over a 3-year span. • Redesigned a 400-node EMR cluster that processes millions of ingestions points reducing the TCO over a 3-year period from $16,800,400 to $3,200,000 in 3 months using a custom logic tree that utilized spot and on-demand instances. • Completed a full-scale rebuild of a complete physical heterogeneous network spanning multiple data centers to a hybrid cloud environment with a single physical origin using GCP and AWS in under 8 months. Reduced TCO by half and increased peak time response rate from 11 minutes to 45 seconds. Completely custom solution using Python, GO, and Terraform. • Fully built out a low latency OCP trading platform management system reducing the deploy time of 300 ephemeral instances from 1 hour to 6 minutes. This allowed our client to reduce their required physical server reserve pool from 40% to 15% reducing the TCO of not only server costs but from loss of power and storage needs. • Designed and implemented the migration from docker-compose to a full k8 cluster on vSphere. This new system reduced the virtual range footprint by nearly half and allowed disparate teams to manage their development ranges. • Responsible for the internal application Development and Operations team which builds and maintains Virtual Cyber Warfare systems for entities such as the DOD, JSOC, and other intelligence agencies around the world.
Canary is the first smart home security device for everyone. We are the 5th largest S3 user in the world in terms of operations at 800,000,000 per day. We ingest around 400 Terabytes of video and run 450,000 hours of video through our CV engine each day. Responsible for all internal engineering teams including ancillary teams such as CV and Data. Inherited and completely revamped an aging and overburdened backend system originally built on Django and MySQL. Migration to Kafka and Cassandra reduced operational complexity and hardened the overall system resiliency. • Migrated the entire engineering organization from waterfall to agile effectively doubling the number of deliverables per sprint. • Spearheaded the cost reduction initiative which has reduced our AWS spend by $3.5 Million dollars a year. Designed and implemented a distributed Nomad cluster for our CV, Ingress, and Notification services reducing operational complexity and reducing our time to scale. • Personally rebuilt the ORM for event and timeline notification engines reducing overall backend load by 60% and decreasing platform downtime from 5% to .001%.