Oscar E. Moreno

Developed payment-processing modules in C, including an ISO8583 card processor. Implemented transaction protocol handling and low-level memory management. Built and maintained GitLab CI/CD pipelines to automate build, test, and deployment for embedded and backend components.

Owned a core risk-assessment service for a global FX & payments platform. Built the majority of the service core logic in Python (FastAPI, SQLAlchemy, Django) and exposed APIs consumed by trading and risk workflows. Configured Kubernetes workloads to maintain responsiveness under variable load, adapting shared infrastructure patterns into the team codebase. Implemented end-to-end observability using Datadog monitors-as-code, alerting, and tagging, plus Sentry error tracking to improve incident detection and response. Ran production workloads on AWS (EC2, RDS) in a microservices environment.

Architected a Rust-based financial orchestration library with modular adapters for multiple LLM providers (OpenAI, Google AI, Anthropic, Deepseek, local). Engineered a gRPC remote-executor pool with smart batching and output caching, doubling inference throughput. Built RAG workflows to parse, embed, and answer natural-language queries over large SEC filings and financial statements. Exposed the Rust core to Python via PyO3 and maturin, shipping a pip-installable package for end-to-end AI finance pipelines. Integrated GPU-accelerated remote compute clusters with robust retry/error handling and CI/CD for deployment.