Senior staff software Engineer and Cloud Architect specializing in distributed storage, systems automation, FinOps, and cloud-native cost optimization.

My journey in open source spans more than two decades—moving from raw kernel-space cluster filesystems on bare metal in the early 2000s to managing highly distributed, multi-million dollar Kubernetes architectures today.

Currently focused on enterprise cloud efficiency, telemetry, and platform engineering at Zendesk, driving the adoption of open standards for cloud financial management.

• CNCF OpenCost Contributor: Actively driving the implementation and optimization of OpenCost at scale.
• Maintainer of opencost-parquet-exporter: Author and maintainer of a specialized data engineering utility that transforms Kubernetes and cloud cost allocation metrics directly into Apache Parquet format, allowing seamlessly integrated big-data analytics and long-term FinOps auditing.
• Cloud Native Architecture: Engineering automated continuous delivery, high-performance monitoring stacks (Prometheus/Grafana), and container orchestration platforms.

Before modern cloud providers abstracted storage under APIs, I designed, deployed, and standardized high-performance, fault-tolerant bare-metal networks.

• Lustre, GFS, & OCFS2: Architected complex storage layouts utilizing Lustre parallel systems, Red Hat Global File System (GFS), and Oracle Cluster File System (OCFS2). Specialized in global lock management and multi-node concurrent environments.
• DRBD Primary/Primary Clusters: Contributed to early architecture discussions and setups (such as the DRBD-user groups) regarding multi-primary node locking mechanisms, split-brain mitigations, and block-level replication using DRBD 0.8.
• Upstream GlusterFS Packaging: Hand-rolled the original standalone "Debianization" of core glusterfs-client and glusterfs-server builds (versions 1.3.x) to enable deployment on Debian/Ubuntu platforms before it moved to modern collaborative code repositories.
• FUSE & Mirroring Utilities: Introduced and managed Debian Requests for Sponsorship (RFS) for utilities like ChironFS—a pioneering, application-layer FUSE replicating filesystem.

• Project Coordinator & Technical Consultant: Served under the Secretaria de Logística e Tecnologia da Informação (SLTI) within the Brazilian Ministry of Planning, Budget, and Management.
• The Blueprint Manuals: Co-authored and coordinated major national technical standards, including:
  • Guia de Administração de Cluster e Grid: The official implementation framework for government high-availability, parallel, and supercomputing grids.
  • Guia Livre: The structural reference framework used to guide the Brazilian Federal Government's historic migration to open-source software backends.

• FISL (Fórum Internacional Software Livre): Regular technical speaker and panelist at the legendary open-source conferences in Porto Alegre (including FISL 8 and FISL 9), debating virtualization, FUSE filesystems, and clustering architecture alongside key community figures like Marcelo.
• National Conferences: Delivered specialized infrastructure workshops and panels at LatinoWare (Foz do Iguaçu) and CONISLI (São Paulo) to advance Linux platform engineering in South America.

[Then: 2003-2010] ---> C, Bash, Linux Kernel, Xen, DRBD, FUSE, Debian Packaging, Bare-Metal SAN/NAS
[Now: 2021-2026] ---> Go, Python, Kubernetes, OpenCost, Apache Parquet, AWS, Prometheus, Terraform, FinOps

• 2003 – 2006: Systems migration, public infrastructure strategy, and consulting for the Brazilian Government.
• 2007 – 2010: Advanced cluster filesystems, open-source Linux storage packaging, and national tech conferences (FISL).
• 2011 – 2020: Enterprise systems engineering, SRE transitions, multi-tenant automation, and cloud migration.
• 2021 – 2026: Advanced Kubernetes infrastructure engineering, container metric pipelines, and open-source cloud FinOps tools.

• GitHub: @lmello