Senior Product Marketing Manager, AI Compute

The Role

You will own go-to-market for Telnyx's AI Compute products: inference, embeddings, AI Gateway and the GPU infrastructure beneath them. Telnyx runs inference on its own NVIDIA GPUs inside its private global network and ships new compute products and models every month. You'll turn that cadence into positioning, launches, public benchmarks and sales enablement that win ML platform teams and voice-AI builders, working day to day with product, engineering, sales and demand generation. This is a senior individual-contributor role with full ownership of a product line, on a marketing team that runs on code and agents rather than SaaS: you'll improve on PMM workflows and always look to automate parts of the job.

What you'll own

You'll own the full AI Compute line: inference (serverless and dedicated GPU), embeddings, AI Gateway and Decision Models. Your buyers are ML platform teams, inference buyers, and voice-AI builders choosing a speech stack. The story you'll tell is that performance is a function of the network and the hardware: Telnyx is carrier-native, runs on its own GPUs, and adds no cloud margin to every token.

You'll prove that story through public latency and quality benchmarks, inference economics and TCO, and competitive campaigns against hosted-inference and speech providers. Success is measured by inference consumption, dedicated GPU contract value, and pipeline attributed to compute launches and campaigns.

The opportunity

• Own positioning and messaging for every AI Compute product, from the model
• catalogue to dedicated GPU contracts, and keep it consistent across web, sales decks,
• docs and campaigns.
• Run launches end to end: tiering, launch brief , product update emails, web pages,
• enablement, and the post-launch read-out on activation and pipeline.
• Lead the benchmarking program as a marketing asset: decide which latency , quality
• and cost comparisons to publish, work with engineering on methodology , and turn
• results into pages, blogs and sales proof points.
• Build the inference economics story: T CO and savings calculators, pricing narrative
• versus hyperscalers and hosted-inference providers, and the competitive battlecards
• behind it.
• P artner with demand generation on compute campaigns (competitive, webinar series,
• paid) by owning the brief , the ICP , the offer and the conversion path.
• Design product-activation and ICP-based lifecycle email for compute users with sales,
• and hand the build to the G TM engineering team.
• Enable sales: battlecards, one-pagers, objection handling, sales labs and the pricing
• and competitor data AEs and BDRs rely on.
• Maintain the competitive map for inference and speech (hosted-inference providers,
• speech vendors, hyperscaler AI services) and feed it into roadmap conversations with
• product.
• Be the compute voice for analyst briefings, customer stories and press.

What we are looking for

Must have

• 5+ years in product marketing for developer-f acing infrastructure, with at least 2 years
• on AI/ML, cloud compute, GPU, or data platform products.
• W orks AI-first, not AI-assisted. Our marketing stack is agents rather than SaaS: you'll
• work with the teams shared agents day to day and continuously build and improve
• autonomous agents for reporting, campaign tracking, ICP research and content
• production.
• Can explain inference serving, GPU economics, latency and throughput trade-offs, and
• speech model quality to an ML engineer without a translator , and to a CFO in one slide.
• Has run launches for a f ast-shipping product line (monthly or f aster) and can show the
• launch system, not just the launches.
• Writes clearly and f ast: pages, briefs, battlecards and talk tracks are the daily output.
• Comfortable owning a number (consumption, pipeline, T CV) jointly with sales.

Nice to have

• Experience marketing usage-based or consumption-priced products.
• Hands-on with an inference API, open-weight models or a speech API; can run a demo
• from the terminal.
• Background at a hosted-inference, speech, GPU cloud company , or marketing against
• them.
• Exposure to analyst relations (Gartner , Frost, G2) and customer-story programs is a
• plus.