Sebastian Sastre

Backend engineer on a Rust microservice for user incentives and rewards at a global crypto exchange: enrollment, eligibility, abuse detection, payouts, and programs that drive activation and revenue.

- Owned several features in the incentives lifecycle system in Rust from scratch. Designed a multi-stage state machine serving hundreds of thousands of enrolled users and well over a billion data points. Built fault tolerance and self-healing into the engines so campaigns ran under my end-to-end ownership without production incidents intentionally by design, not by incident-driven patches.

- Rebuilt payment processing as an idempotent event-driven engine. Replaced a table-based implementation with a generic, scalable pipeline that could serve payment sources beyond the original use case, for hundreds of thousands of users.

- Built an AI-assisted, spec-driven delivery harness with review gates (not ad-hoc prompting). Shipped 10+ production features under that workflow; in one week, delivered four in parallel in the same Rust codebase.

- Partnered with product, peer engineering teams, and marketing technology on requirements and on downstream consumers of events from owned features.

- Stack: Rust · event-driven services · Kafka-style pipelines · PostgreSQL

Designed, tested, and shipped a high-performance Rust caching system to extend main backend capability beyond 25K operations per second (ops), directly facilitating the seamless expansion of core production roaming infrastructure to over 1 PiB of data.

• Led a cross-functional team of 4 engineers through a massive 6-month legacy code refactoring initiative to resolve deep technical debt, leveraging advanced Bash, Python, and Docker Swarm to achieve full containerization readiness.
• Engineered a custom, multithreaded stress-testing framework in Go to inject high-volume synthetic traffic into the production pipeline, guaranteeing strict enterprise SLA preservation.
• Supervised large-scale persistence migrations (Voyage to Mapless) and maintained high-availability data infrastructure with a meticulous focus on complex PostgreSQL replication structures.

Led a team of 6 engineers to architect and ship a next-generation, high-volume ad server layer capable of routing massive globally distributed traffic.

• Designed and deployed automated, backward-compatible infrastructure pipelines using GCloud, enabling real-time, zero-downtime global multi-region application deployment.