Director of AI Systems Solutions Engineering
Posted Updated
Description
- Tensordyne is hiring a Director of AI Systems Solutions Engineering to own and grow our most important technical customer engagements
- This is a senior, highly technical role for someone who understands modern AI infrastructure from model architecture through accelerator hardware, distributed inference, serving software, and datacenter deployment — and who can credibly engage with the engineers and architects building the next generation of AI platforms
- You will work directly with frontier model builders, hyperscalers, neoclouds, developers, infrastructure partners, and strategic customers as they evaluate and deploy Tensordyne systems
- You will also build and lead a small team of exceptional Sales and Solutions Engineers responsible for customer benchmarking, technical evaluation, NPI, model enablement, AI DC architecture, and production deployment
- This is not a traditional pre-sales engineering role. The team will operate at the frontier of a rapidly changing technology landscape, working with constantly evolving new models and requirements. The right person will be equally comfortable in a customer architecture review, helping prioritize product capabilities and roadmaps, and leading a technical evaluation with hyperscalers and frontier AI companies
- Strategic technical customer engagements: Own the technical relationship with key customers and partners from initial architecture discussions through benchmarking, evaluation, integration, deployment, and expansion
- Technical evaluation strategy: Define how Tensordyne demonstrates system performance across KPI’s like throughput, tokens/sec/user, ttft, memory utilization, power efficiency, system density, model accuracy/quality, and other relevant inference metrics
- AI workload and model architecture engagement: Work with customers and model developers to understand current and emerging HW and model architectures, serving requirements, context lengths, parallelism strategies, model topology, quantization approaches, and inference optimization requirements
- Benchmarking and competitive analysis: Maintain a technically rigorous understanding of Tensordyne performance relative to leading GPU and AI accelerator platforms. Ensure customer-facing comparisons are credible, reproducible, current, and aligned with real deployment requirements
- Model enablement and optimization: Partner with compiler, runtime, kernel, systems, and SDK teams to bring important customer models and workloads onto the Tensordyne platform and identify opportunities for performance improvement
- Forward deployment: Lead technical PoCs, remote evaluations, on-premises beta deployments, integration programs, and production readiness efforts with strategic customers
- Customer-to-product feedback loop: Translate recurring customer requirements into clear priorities for the SDK, compiler, runtime, inference server, model support, orchestration, networking, observability, and system architecture teams
- Technical market intelligence: Stay deeply current on model architectures, inference techniques, accelerator roadmaps, serving frameworks, competitive systems, benchmarking methodologies, and changes in the AI infrastructure market
- Team leadership: Help recruit, develop, and lead a small team of elite Sales/Solutions Engineers capable of independently managing sophisticated technical engagements with the world’s most demanding AI infrastructure customers
- Scalable technical GTM: Turn early customer engagements into repeatable benchmarks, evaluation frameworks, reference architectures, deployment playbooks, documentation, demos, and technical collateral that can support a rapidly growing customer base
- Tensordyne is entering the stage where our technology moves from internal development into the hands of customers
- The technical customer organization will play a central role in that transition
- This team will help determine which workloads we prioritize, how customers evaluate our platform, how quickly new models become production-ready, how effectively customer feedback reaches engineering, and ultimately how Tensordyne systems are adopted at scale
- We are looking for someone who wants to build that capability from the beginning — and establish the technical standard for how Tensordyne engages with the companies defining the future of AI infrastructure
- We are looking for a proven AI systems leader with substantial technical depth and strong judgment
- Strong knowledge of AI accelerator and system architecture, including compute, memory hierarchy, interconnect, parallelism, and distributed inference
- Deep understanding of modern AI inference systems, including LLM and multimodal architectures
- Demonstrated ability to engage technically sophisticated external organizations including Sr technical leaders at hyperscalers, cloud providers, model developers, AI infrastructure companies, or large enterprise engineering teams
- Ability to operate effectively in a fast-moving environment where the product, software stack, competitive landscape, and customer requirements are evolving simultaneously
- Experience benchmarking and optimizing workloads on large-scale AI infrastructure
- Experience reasoning about inference performance across latency, throughput, memory bandwidth, utilization, batching, context length, prefill, decode, and system scaling
- Strong understanding of production inference techniques including quantization, tensor/model/expert parallelism, disaggregated serving, KV-cache management, distributed execution, and related optimization strategies
- Hands-on familiarity with modern AI frameworks and serving environments such as PyTorch, vLLM, SGLang, Triton, or comparable systems
- Experience working across the boundary between AI software and accelerator hardware, ideally including GPUs, custom silicon, or emerging AI accelerators
- Experience leading high-performing Solutions Engineering, Field Engineering, Forward Deployed Engineering, or comparable technical customer teams
- Large-scale AI inference infrastructure
- High-performance computing or distributed systems
- Experience bringing a new accelerator architecture or AI infrastructure platform from early access through customer validation and production deployment is especially valuable
- GPU or custom AI accelerator platforms
- Frontier or foundation models
- Candidates may come from organizations building or deploying:
- Hyperscale cloud infrastructure
- AI serving and orchestration platforms
- Compiler, runtime, or distributed AI systems
Skills
Sign in to read and join the discussion.