Saul Reyna

Go, PostgreSQL, Kubernetes, GCP, AWS

• Designed and shipped a policy-driven data lifecycle system for multi-tenant PostgreSQL. Retention runs through batched, timeout-bounded procedures with advisory locking and dynamic batch sizing, and is rolled out across 6+ production regions with minimal impact on serving traffic.
• Hardened a low-latency PostgreSQL wire-protocol proxy with per-connection statement and idle-transaction timeouts, structured error responses, and replica-aware routing, eliminating a class of unbounded-query incidents and reducing on-call load.
• Instrumented both systems end to end with metrics, dashboards, and alerting, treating operability as a first-class requirement.
• Extended a TypeSpec code-generation toolchain that emits Go structs and PostgreSQL DDL, keeping services and database schemas consistent by construction.

Distributed systems, data pipelines, ML platforms

• Built and operated distributed microservices powering AI-based sales and marketing products, designed for reliability and performance at enterprise scale.
• Designed large-scale ingestion and batch pipelines feeding ML-driven features, with monitoring, alerting, and self-healing that reduced manual operational work.
• Partnered with AI researchers and ML engineers to take research ideas to production, building data platforms and generative-AI tooling.

Maintain and tune the archive's on-prem WordPress site, and built import workflows that let data-entry volunteers add records faster.