15+ years in network engineering and interconnection at internet, CDN, cloud, or AI-infrastructure scale.
Production accountability for edge connectivity and/or backbone optical transport supporting large-scale GPU or AI infrastructure.
Hands-on experience with InfiniBand and RoCEv2 GPU fabrics, including migration planning and tuning.
Deep expertise in peering and IX, with a proven track record in network economics.
Strong skills in DCI and long-haul design, including 400G-class DCI and optical transport.
WAN-scale routing expertise in eBGP/iBGP, MPLS/RSVP-TE, and EVPN-VXLAN.
Bachelor's degree in Computer Science, Electrical Engineering, Telecommunications, or equivalent experience.
Responsibilities
Own the traffic-exchange architecture and cost model for AI cloud and data center sites.
Define and lead peering and IX strategy, representing Bitdeer in the community.
Set physical design standards for meet-me rooms and border connectivity.
Design hyperscaler interconnects and agree on burst profiles with partners.
Lead build-versus-lease architecture and costing for DCI and metro transport.
Model GPU-cluster SLA exposure and translate it into design requirements.
Define InfiniBand and RoCE posture to support GPU backend fabrics.
Benefits
A culture that values authenticity and diverse perspectives.
An inclusive, respectful environment with open workspaces.
Opportunities to network with industry pioneers and enthusiasts.
Direct contribution to the future of the digital asset and AI computing industry.
Involvement in new projects and developing processes/systems.
Personal accountability, autonomy, and rapid growth opportunities.
Attractive benefits including training and mentoring.
Full Job Description
We are seeking
We are seeking a Senior Director, Interconnection & Network Economics to own the architecture and economics of how Bitdeer's AI cloud connects to the rest of the world. You will decide what our border, interconnection and long-haul connectivity should be and what it should cost, from peering and Internet Exchange (IX) strategy and hyperscaler on-ramps to build-versus-lease decisions for data center interconnect (DCI). You will turn GPU-cluster performance requirements into a north-south network design and a delivered cost curve. You will work alongside our Network Infrastructure & Connectivity team, which sources, contracts and operates that connectivity, and set the target architecture and cost model they deliver against. This role suits a recognized interconnection practitioner who has run edge connectivity and optical transport for production GPU infrastructure.
What you will be responsible for
Own Bitdeer's traffic-exchange architecture and its delivered cost curve ($/Mbps, $/Gbps), and set the target interconnection architecture and cost model for our AI cloud and data center sites.
Define and lead peering and IX strategy, including Bitdeer's public and private interconnection posture and IX participation, and represent Bitdeer in the peering and IX community.
Set meet-me room (MMR) and border physical design standards, including separated MDFs, distinct power and cooling failure domains, and dual-homed cluster border switches.
Design hyperscaler interconnect, including AWS Direct Connect, Azure ExpressRoute and Google Cloud Interconnect, and agree in advance the burst profiles both sides must support.
Lead build-versus-lease architecture and costing for DCI, metro transport and dark fiber, and recommend the position for leadership approval before it is sourced and delivered.
Model GPU-cluster SLA exposure at the border, including path-length asymmetry, ECMP hashing, jitter and cluster dropout, and turn it into design requirements and SLA-credit risk.
Define Bitdeer's InfiniBand and RoCE posture at the north-south boundary so that border design supports GPU backend fabrics and their migration path.
Contribute production InfiniBand and RoCEv2 experience to GPU cluster fabric design reviews (fat-tree, rail-optimized and dragonfly topologies), so that east-west fabric choices and the north-south border are designed together.
Advise on fabric performance standards, including adaptive routing, DCQCN/ECN congestion control, traffic isolation and NCCL behavior, where they affect border design, cluster SLAs or cost.
Select border platforms, including router versus switch and FIB and hardware scale, and recommend standards for approval and deployment.
Advise on connectivity and GPU fabric incidents that have an architectural or cross-boundary root cause, and feed the lessons back into design standards and the cost model.
How you will stand out
15+ years in network engineering and interconnection at internet, CDN, cloud or AI-infrastructure scale.
Production accountability for edge connectivity and/or backbone optical transport supporting large-scale GPU or AI infrastructure across multiple data centers.
Hands-on production experience with InfiniBand and RoCEv2 GPU fabrics (fat-tree, rail-optimized or dragonfly topologies; Nvidia, Arista and Cisco platforms), including RoCE migration planning, and familiarity with UFM, adaptive routing, DCQCN/ECN congestion control and NCCL tuning.
Deep peering and IX expertise and a track record in network economics: measurably reducing delivered cost per Mbps while supporting traffic growth across large IX footprints and multi-Tbps peak traffic.
Strong DCI and long-haul design skills: 400G-class DCI, DWDM and optical transport, build-versus-lease costing, optical lead times and in-line amplifier (ILA) site support.
WAN-scale routing expertise in eBGP/iBGP, MPLS/RSVP-TE and EVPN-VXLAN, with an informed position on SRv6.
Ability to connect network behavior to business outcomes, such as tracing jitter through ECMP hashing to GPU job dropout and SLA-credit exposure.
Clear written and verbal communication, with the ability to make the architecture and build-versus-lease case to senior leadership and to work closely with procurement and operations teams.
Bachelor's degree in Computer Science, Electrical Engineering, Telecommunications or a related field, or equivalent industry experience.
Recognized standing in the interconnection community
What you will experience working with us
A culture that values authenticity and diverse perspectives.
An inclusive, respectful environment with open workspaces and an energetic, start-up spirit.
A fast-growing company with opportunities to network with industry pioneers and enthusiasts.
The ability to contribute directly and make an impact on the future of the digital asset and AI computing industry.
Involvement in new projects and in developing processes/systems.
Personal accountability, autonomy, rapid growth, and learning opportunities.
Attractive benefits and development such as training and mentoring.
