Senior Software Engineer, AI Platform Reliability

Built a Claude-driven code review assistant for trading-support repositories, reducing review turnaround from two days to four hours across more than 30 weekly pull requests. Refactored prototype agent scaffolding into a production service supporting tool and function calling against PostgreSQL, with prompt engineering and model evaluation suites to detect regressions before deployment.

Owned the reliability layer for agent workloads, including Google Cloud tracing and logging, Grafana dashboards, database permissions, connection pooling, and migration reviews across six internal schemas. Implemented GitHub Actions release controls and Customer.io event triggers, delivered more than 15 reliability fixes, and advocated for Kubernetes CronJobs over a custom scheduler.

Operated observability and stability for an internal platform used by more than 200 researchers. Instrumented tracing and alerting that identified approximately three regressions per quarter before users submitted support tickets.

Containerized legacy build tooling in Kubernetes, reducing new-engineer environment setup from two days to 30 minutes. Migrated batch reporting to Google Cloud, reduced a nightly job from four hours to 25 minutes, automated deployments with GitHub Actions, maintained an artifact registry, and mentored four junior engineers.

Delivered custom application development across 12 client engagements, taking features from early scoping through implementation, testing, and release. Worked across varied client requirements while owning end-to-end delivery responsibilities.

Introduced QA automation to replace a manual regression process with a repeatable test suite. Refactored a legacy billing integration, eliminating 40% of duplicated code and reducing support tickets related to that module.

Built scalable backend services in a cross-functional product and infrastructure environment, supporting multiple internal teams. Collaborated on shared systems with an emphasis on dependable data access and service quality.

Increased test coverage for a shared data-access layer from 55% to 80%, helping reduce rollback incidents. Authored technical design documents reviewed by staff engineers and adopted by two sibling teams.

Developed components of an electronic trading platform, with a focus on order-routing reliability and production logging. Supported systems where operational correctness and timely diagnostics were essential.

Improved the monitoring stack through threshold tuning and clearer dashboards. Reduced false alerts by 35%, improving signal quality for engineers responsible for production response.

Bachelor of Science and Engineering in Computer Science.