Cloudflare adds two new infrastructure leaders

Cloudflare has announced two senior hires to lead its global infrastructure and hardware efforts. Marcelo Affonso joins as VP of Infrastructure Operations, and Rebecca Weekly comes on as VP of Hardware Systems. Both bring deep experience in scaling large cloud and computing operations.

Marcelo Affonso: expanding network reach and operations

Affonso will be responsible for building, expanding, and optimizing Cloudflare’s global network footprint. His background spans logistics, supply chain, and cloud infrastructure, including nearly a decade at Amazon where he managed fulfillment operations across the UK, Italy, and Canada before moving to AWS to support one of its largest data center regions.

In announcing his move, Affonso emphasized the need to make the Internet more secure, reliable, and private for a wider range of users, not just large enterprises. He also noted the importance of adapting approaches across markets — a perspective shaped by living and working in seven countries.

Cloudflare operates a network within 50 milliseconds of 95% of the global Internet-connected population. Affonso said he was drawn to the company's pace of innovation and its mission to democratize access to key technologies. Recent infrastructure milestones include adding 18 new cities to the network, bringing coverage to over 270 cities globally, and expanding Cloudflare Network Interconnect (CNI) locations. Affonso also highlighted planned work on Cloudflare R2 Storage and Cloudflare for Offices, which connect office traffic directly to Cloudflare’s network.

Rebecca Weekly: bringing domain-specific hardware design to the edge

Weekly comes to Cloudflare with a career in semiconductors and system design, spanning graphics chips, ASIC validation, and CPU design for hyperscale operations. She was previously Chairperson of the Open Compute Project Foundation, where she focused on driving hyperscale innovation from the cloud to the edge.

In her new role, Weekly will design hardware systems to support Cloudflare’s network. She points to what John Hennessy and David Patterson called the “new golden age of computer architecture” as a key motivation. Hennessy and Patterson argue that with the end of Dennard scaling and slowing Moore’s Law, the industry needs to address inefficiencies in general-purpose x86 and ARM processors. Their roadmap highlights three areas:

  • Optimizing high-level language performance on existing infrastructure, which has historically favored developer efficiency
  • Domain-specific architectures that optimize hardware parallelism for a particular computational use case
  • A hybrid approach where domain-specific languages enable domain-specific architectures for broader infrastructure gains

Weekly said she wanted to work more closely with application developers to achieve architectural optimizations that neither hardware nor software teams could accomplish separately. Cloudflare combines deep software services with control over the hardware running its global network, a model that has scaled to handle an average of over 32 million HTTP requests per second — a significant share of Internet traffic.

The company plans to continue sharing its hardware findings with the broader industry through initiatives like the Open Compute Project. Both new hires underscore Cloudflare’s focus on making its infrastructure operations and hardware design more efficient as it expands its network worldwide.