Principal Product Manager

Trust In SODA; Greater London; On-site; GBP 100,000 - 200,000; Full time

Benefits offered by this job

25 days leave

Private medical insurance

Gympass

Cycle to work

Company shares programme

Enhanced parental pay and leave

Job summary

Trust in SODA is seeking a Principal Product Manager, Operations (AI infrastructure) to own the hardware/infrastructure lifecycle from data centre kit arrival to running in the platform. You will coordinate across support, DC ops and platform teams, building internal tooling and self-service capabilities.

You’ll have an engineering background with strong API, Kubernetes, cloud, automation and security chops, and be comfortable guiding design discussions while staying close to hands-on work.

Qualifications

• Expert in infrastructure operations with hands-on hardware/datacentre experience.
• Strong familiarity with APIs, Kubernetes, cloud, automation, observability and security.
• Proven ability to pull support, DC ops and platform teams together to fix or ship.

Responsibilities

• Own strategy, roadmap and delivery for internal tooling, CMDB, service catalogues and operational workflows.
• Set up reliable sources of truth for infrastructure, services, customers, environments and configuration.
• Collaborate with SRE, infrastructure, security and support to reduce operational friction and enable self-service.
• Make build vs buy calls on third-party tools.
• Work alongside engineers, challenge architecture and understand how things are built.

Skills

Infrastructure ops

Kubernetes

Cloud

APIs

Automation

Observability

Security

Tech design

Tools

NetBox

ServiceNow

Terraform

Grafana

Job description

Principal Product Manager, Operations (AI infrastructure)

This AI infrastructure business is rebuilding its product team around technical PMs who each own an engineering track, and they'd rather hire an engineer and coach the product side than the other way round.

About the company

A startup running dense, large-scale GPU and CPU compute, including the latest NVIDIA hardware, for teams training, deploying and scaling AI models. Product is moving from generalist PMs to a vertical model where each PM owns a dedicated engineering track. It moves fast and favours speed over structure.

The role

You'll own the product side of the hardware and infrastructure lifecycle, from new kit arriving in a data centre to it running in the platform. A lot of the job is making that path work across support, customer, data centre ops and platform teams, and building the internal tooling behind it.

• Own strategy, roadmap and delivery for internal tooling, CMDB, service catalogues and operational workflows
• Set up reliable sources of truth for infrastructure, services, customers, environments and configuration
• Work with SRE, infrastructure, security and support to cut operational friction and give teams more self-service
• Make build vs buy calls on third-party tools
• Sit alongside your engineers, challenge architecture and understand how things are built

What we're looking for

You'll be a subject matter expert in infrastructure operations rather than a generalist PM. That usually means time somewhere that runs a lot of its own hardware: a hyperscaler, a hedge fund, an exchange or any business with a large on-prem estate. AI experience isn't needed.

An engineering background counts for more than PM credentials. You'll be comfortable with APIs, Kubernetes and cloud, inventory and dependency data, automation, observability and security, and able to hold your own in technical design discussions.

You'll have pulled support, customer, DC ops and platform teams together to get things fixed or shipped, and you'll enjoy ambiguity and speed. This is a principal role that stays close to the technical detail, so it won't suit someone who has moved well away from hands-on work. If you've stepped back from a Head or Lead role and want to build a function from scratch, that's welcome.

Nice to have: NetBox, ServiceNow, Terraform, Grafana or similar, and a good grasp of distributed systems failure handling (retries, idempotency, eventual consistency).

Salary and benefits

Base from £100,000, with the total package into the £200,000 range once bonus and backdated equity are included. There's flexibility for the right person. Benefits include 25 days' leave, dedicated learning time, private medical insurance, a Gympass subscription, cycle to work, a company shares programme and enhanced parental pay and leave.