Technical Lead, Flashblade//EXA

Santa Clara · USD 239,000-359,000 / year · Onsite · Posted Oct 8, 2026

At a glance

As the Performance Technical Lead for FlashBlade//EXA, you will define and own the end-to-end performance narrative for Everpure’s exascale disaggregated storage architecture built for AI factories.
In this strategic and hands-on role, you will bridge systems performance engineering, benchmark architecture, and competitive intelligence to establish how //EXA is engineered, proven, and positioned against leading competitors in HPC and hyperscale AI infrastructure.

Architect and Define Performance Roadmaps: Translate deep storage protocol expertise (NFSv3 direct-to-DN, pNFS, S3) into technical performance requirements for //EXA’s disaggregated data path to resolve GPU-starved read patterns, metadata storms, and bursty HPC workloads.
Drive Strategic Market Alignment for GPU/Sovereign Cloud: Align //EXA’s performance roadmap, sizing guidance, and reference architectures with the buying criteria and operational needs of multi-tenant GPU-cloud and sovereign-cloud operators.
Establish Competitive Performance Technical Positioning: Develop rigorously substantiated architectural analyses comparing metadata scaling, durability tradeoffs, and GPU-direct paths to empower field engineering during customer technical evaluations and POCs.
Optimize Stack-Wide Tuning and Benchmarking: Lead end-to-end performance validation across storage, RDMA networking, and compute, while building reproducible benchmark suites (MLPerf Storage, IO500, elbencho) that validate public claims and demonstrate market superiority.
Define Multi-Tenant QoS and SLA Guarantees: Establish technical requirements and validation frameworks for quality-of-service, workload isolation, and IOPS limits to deliver predictable SLA commitments in shared deployments.

Storage Performance Architecture Expertise: Proven expertise in storage performance engineering across file, block, and object protocols (such as parallel filesystems, pNFS/NFS at scale, and S3-scale object stores).
GPU Cluster & Fabric Knowledge: Strong working knowledge of GPU cluster architecture, including RDMA fabrics, GPUDirect Storage, checkpoint/restore dynamics, and resolving bottleneck patterns across compute, network, and storage layers.
Benchmarking & Load Generation Mastery: Proficiency with industry benchmark standards (MLPerf Storage, IO500) and load-generation tools (elbencho, fio, vdbench), alongside the capability to design custom, workload-representative tests.
Competitive Analysis & Systems Design: Demonstrated ability to perform architecture-level competitive analysis, along with a firm grasp of multi-tenant resource management concepts (QoS, rate limiting, and workload isolation) in distributed systems.
Cross-Functional Technical Leadership: Ability to navigate complex technical discussions across the full stack—from debugging low-level network queue stalls to articulating system capabilities directly to customer evaluation teams.

We are primarily an in-office environment and therefore, you will be expected to work from the Santa Clara, CA office in compliance with Everpure’s policies, unless you are on PTO, or work travel, or other approved leave.