I

Inferact

Open-source LLM inference engine with scalability

Technical Product Marketing Manager

Full-TimeUpdated on 10/14/2026
No salary listed
Mid
Bachelor's
San Francisco, CA, USA
In Person
Growth & MarketingProduct Marketing
H1B Sponsorship Available

About the job

Requirements
  • A bachelor's degree in Computer Science, Mathematics, Engineering, or a related technical discipline, or equivalent experience as an engineer or developer.
  • Experience owning product marketing or developer relations for open-source software, LLM infrastructure, AI infrastructure, or developer tools.
  • Strong product marketing fundamentals across positioning, messaging, audience definition, narrative development, launches, and practical go-to-market execution.
  • Ability to write and edit technical launch content credible with developers, machine learning engineers, and infrastructure engineers.
  • Experience working cross-functionally with engineering, product, leadership, design, sales, and external partners.
  • A hands-on execution mindset and ability to build launch systems, templates, and playbooks in a fast-moving startup environment.
Responsibilities
  • Own product marketing for vLLM and Inferact across launches, positioning, technical narratives, partner announcements, and sales enablement.
  • Own day-zero launch communications, product positioning, the benchmark narrative, customer proof, and the forward product marketing roadmap.
  • Become the single owner for launch communications across vLLM and Inferact, including blog content, social copy, partner coordination, and the internal readiness checklist.
  • Ship at least two launches as primary author and publish a launch playbook that engineers and partners can follow without additional direction.
  • Build a competitive teardown of recent launches and use it to establish the quality bar for future launches.
  • Own positioning and messaging, including personas, competitive talk tracks, and website product pages.
  • Own partnership announcements with ecosystem partners such as Google, AMD, and NVIDIA from messaging through launch day.
  • Own the sales collateral kit and standard vLLM and Inferact talk template, and run enablement for engineers presenting externally.
  • Own product messaging across vLLM and Inferact, including technical product breakdowns, benefits, customer case studies, and supporting go-to-market materials in close partnership with go-to-market teams.
  • Present a 2027 product marketing plan covering launch tiers, the benchmark program, case study pipeline, and partner relationships.
  • Bring market and customer input into product roadmap and pricing and packaging discussions.
  • Create a structured feedback loop between Marketing and Sales with Product and Engineering, ensuring customer, market, and go-to-market feedback consistently flows into product and launch decisions.
  • Build and operationalize a repeatable marketing process for launching vLLM and Inferact content, covering intake, messaging, review, distribution, partner amplification, and post-launch measurement.
Desired Qualifications
  • Deep familiarity with AI infrastructure, model serving, vLLM, open-source infrastructure, cloud platforms, or GPU ecosystems.
  • Experience owning benchmark narratives, technical product pages, customer case studies, or developer-facing sales enablement.
  • Experience leading joint go-to-market efforts with cloud providers, hardware vendors, model companies, or open-source ecosystem partners.
  • A track record of building product marketing programs from 0 to 1 and turning reactive launch work into a proactive roadmap.
  • Direct experience marketing inference, model serving, or other performance-sensitive AI infrastructure.
  • Experience launching open-source projects or commercial products built around an open-source ecosystem.
  • Ability to create lightweight technical assets, landing pages, demos, or workflows using modern AI and no-code tools.

About the company

I

Inferact

View
Website

Inferact builds AI inference infrastructure by maintaining vLLM, an open-source LLM inference engine, and offering a managed enterprise inference service. vLLM uses PagedAttention to optimize GPU memory, cutting inference costs and latency while preserving model quality. The company supports multiple architectures and hardware, aligns with PyTorch Foundation governance, and pursues open-source collaboration alongside a commercial platform. Its goal is to turn AI inference into a reliable, scalable operating layer of the AI stack, separating model deployment from application development.

Company Size

11-50

Company Stage

Seed

Total Funding

$150M

Headquarters

San Francisco, California

Founded

2025

Get referred to Inferact

See people who can refer or advise you

Simplify Jobs

Simplify's Take

What believers are saying

  • Inferact raised $150 million in January 2026 at an $800 million valuation.
  • TPU megakernel work reached 709 tokens per second on Kimi K3.
  • Open-source leadership keeps contributors flowing from Meta, Google, and Amazon ecosystems.

What critics are saying

  • NVIDIA Dynamo commoditizes vLLM, SGLang, and TensorRT-LLM beneath NVIDIA-controlled routing.
  • RadixArk and ZML target the same inference layer, compressing pricing power.
  • If managed inference lags open-source innovation, customers keep vLLM and skip Inferact.

What makes Inferact unique

  • Inferact owns vLLM, the default inference engine for 500-plus architectures.
  • Google partnered with Inferact on TPU vLLM support on September 14, 2026.
  • Inferact ships day-zero quantized releases like Qwen3.8-2.4T-A95B with NVIDIA and AMD.

Help us improve and share your feedback! Did you find this helpful?

Benefits

  • Health Insurance
  • Dental Insurance
  • Vision Insurance
  • 401(k) Company Match

Growth & Insights and Company News

Headcount

6 month growth

6%

1 year growth

10%

2 year growth

10%
vLLM
Aug 12th, 2026
Day 0 support for Qwen3.8-2.4T-A95B on vLLM.

Day 0 support for Qwen3.8-2.4T-A95B on vLLM. vLLM Team and Inferact vLLM is announcing Day-0 vLLM support for Qwen3.8-2.4T-A95B. This is the first model from the Qwen family to bring a Qwen-Max-class model to open-weight release. Qwen3.8-2.4T-A95B is built on the Qwen 3.5 architecture and runs on vLLM out of the box. In addition to the official FP8 and BF16 checkpoints, Inferact has released MXFP4 and NVFP4-quantized weights that match full-precision quality while significantly reducing memory and bandwidth overhead. Qwen3.8-2.4T-A95B is a 2.4-trillion-parameter sparse MoE model featuring 512 experts. Within its 92-layer hybrid backbone, full attention is applied at every 4th layer while the remaining 69 layers run linear attention. As one of the largest open-weight models released to date, running inference requires at least two NVIDIA B300 / AMD MI355X nodes (or a single node for the FP4 quantized version). Tl;dr. * Day-0 support: Qwen3.8-2.4T-A95B reuses the Qwen 3.5 architecture and runs on vLLM from day one with no architecture changes required. * Flexible precision: FP8, BF16, NVFP4, and MXFP4 checkpoints are available. * Multi-vendor optimization: Validated across hardware partners including NVIDIA and AMD. Quick start. For NVFP4: For MXFP4: See the vLLM recipes for the full serving guide and recommended flags. FP4 quantization: quality at lower cost. To minimize inference costs and maximize GPU memory efficiency, the Inferact team quantized selected layers - including the routed experts - to FP4 weights using Round-to-Nearest (RTN) quantization with activation calibration to enable 4-bit activations. vLLM ran initial verifications to confirm that quantization accuracy remains intact. Note that increasing the reasoning budget is required to reproduce these evaluation results. | Benchmark | FP8 | NVFP4 | | GSM8K (strict / flexible) | 89.61% / 90.52% | 90.37% / 91.05% | | AIME25 @3 (avg / pass) | 87.78% / 93.33% | 92.22% / 96.67% | Optimizations. To enable efficient inference for this 2.4T parameter model, vLLM collaborated closely with NVIDIA and AMD to develop optimized kernels based on existing Qwen 3.5 support. On NVIDIA platforms, NVIDIA and Inferact co-developed ultra-fast kernels for Linear Attention (Gated Delta Rule), Attention (GQA), Dense GEMMs, and MoE routing. New fused kernels were added to reduce communication overhead. Significant effort was also dedicated to identifying the best decomposition of work to maximize performance, including combining Data Parallelism and Tensor Parallelism for Attention and Expert Parallelism for the MoE. On AMD Instinct GPUs, vLLM accelerates Qwen3.8 with AITER-fused Gated DeltaNet decode, attention, and MoE kernels, reducing kernel-launch and data-movement overhead. For Shared Expert MoE, the shared-expert path leverages highly optimized hipBLASLt GEMM kernels, while routed experts use AITER FusedMoE. AMD Quark quantization support enables efficient MXFP4 deployment, substantially reducing model memory requirements while maintaining strong accuracy. Deployment tips. The Qwen 3.8 model card recommends the following generation parameters for optimal performance: temperature=1.0, top_p=0.95, top_k=20, min_p=0.0, presence_penalty=0.0, repetition_penalty=1.0 Here is a Python client snippet to query the model once the vLLM server is running. Because Qwen 3.8 is a reasoning model, ensure you allocate a sufficient token budget for agentic workflows by setting a high max_tokens value. Acknowledgements. vLLM thank the Qwen team for releasing the model weights as well as their ongoing collaboration, and its hardware partners, NVIDIA and AMD, for their joint engineering contributions. vLLM also thank the Inferact team for delivering quantized checkpoints and end-to-end vLLM integration, as well as the broader vLLM community. Thanks for its inference partners, including DigitalOcean, Together AI, who helped with early testing.

TechCrunch
Jan 22nd, 2026
Inference startup Inferact lands $150M to commercialize vLLM | TechCrunch

The seed round values the newly formed startup at $800 million.