EU remote
Solutions Engineer – Enterprise Compute
About this role
Location: Remote Employment: Full-Time Start Date: ASAP Experience: 5+ years Language: Fluent English required | French is a PLUS Industry: Cloud Infrastructure / GPU Compute / AI Infrastructure ABOUT THE ROLE You are the technical owner of our B2B customers, from the first call to production and every month after. You run discovery, design the setup, prove it in a POC, onboard it and keep it healthy. At its core this is a pre-sales and technical account management role.
You know our stack well enough toanswer most customer questions yourself, and you bring in our engineers only for the deep end, with thecontext and first diagnosis already done.This is a new, hands-on individual contributor role. No quota. You work side by side with sales and reportto the Compute product lead. Today the product lead covers this job on top of product, while newcustomer channels open at three times the pace of earlier this year.
Responsibilities What you'll do Pre-sales - Run technical discovery with customers' engineers: workload, models, serving stack, latency and throughput targets, storage, network, security and data-residency constraints. - Turn that into the right setup and size it: model to GPUs, CPU, RAM, NVMe and network per server, and bare metal vs full-server VM vs instances vs managed inference. - Answer technical due diligence: bill of materials, PCIe topology, NICs and network design,out-of-band access, data-centre power and connectivity, security and compliancequestionnaires.
You answer most of it yourself, with infrastructure and security behind you forthe deep end. - Represent us in enterprise and public-sector workshops, mapping our stack against theirs: Kubernetes, identity, logging, storage APIs, integration. - Check every technical commitment before it reaches a contract: SLA, performance minimums, acceptance tests, dates for roadmap features. Proofs of concept - Design each POC with written success criteria, a timebox, reserved capacity and a data-handling plan.
- Run the benchmarks with the customer, using standard tools (NCCL tests, fio, vLLM or SGLang load tests) or their own method. Write up the results and the go or no-go. Onboarding - Own technical onboarding end to end: accounts and access, private networking, public IPs and firewalling, load balancing, storage, server handover against an acceptance checklist, first workload live. - Work with infrastructure on new server batches, from burn-in to handover.