Platform Engineer

About Defyner

Marketing is being rebuilt.

For decades, great teams have been constrained by the coordination tax across tools, agencies, markets, and formats. AI changes the production capacity of a marketing organisation, but without governance it mostly creates more activity, more vendor dependency, and more risk.

Defyner exists to fix that. We're the brand engineering company that helps the world's most ambitious brands design, develop and deploy production-grade, governed multi-agent systems inside their own infrastructure, so strategy turns into repeatable, measurable execution at software speed and scale.

The differentiator is no longer the tools you use, but the system that uses them. We call that system the brand harness: codified workflows, deterministic checks where possible, human approval gates, evaluation harnesses, observability, and end-to-end auditability, so enterprises can scale speed without increasing brand risk.

We believe a new craft is emerging: Brand Engineering, translating brand and business logic into machine-usable guidance and execution. The infrastructure to do this reliably at enterprise scale doesn't yet exist, so we're building it.

Why this role now

We’re excited by our strong early momentum with leading marketers across the US, UK and EU.

To meet that demand and accelerate what Defyner can ship, we've raised our angel round and are building a founding team that can deliver world-class work for enterprise brands.

That means moving from early one-off deployments to a repeatable platform layer we can roll out across multiple brands: secure, observable, and reliable by default.

This is an early role in our platform engineering function, at the heart of that shift. You'll help build the shared platform and reliability machinery that makes every deployment safe, controllable, and scalable.

You'll step into a live delivery pipeline spanning enterprise brands across the US, UK, and EU, with real engagements and real infrastructure problems to solve from day one.

What you’ll work on

• The shared platform layer used across deployments: integrations, permissions, audit logging, evaluation, monitoring, and deployment workflows.
• Core patterns for agent reliability: tool contracts, retries, idempotency, human-in-the-loop approval gates, and safe execution.
• A reusable “client stack” approach: deploy inside client infra (security/ownership constraints), with clear separation of client-specific vs reusable components.
• Observability: tracing, metrics, prompt/version tracking, model cost/performance visibility, and incident debugging.
• Quality and safety: evaluation harnesses, regression tests, red teaming patterns, and release processes.

What you’ll do (examples)

• Build an integrations framework (MCP/tooling) that can connect to enterprise systems (CRM, CMS, analytics, ad platforms, project management tools, data warehouses) with permissioning and audit trails.
• Implement an evaluation pipeline that continuously checks agent outputs against brand/compliance rules and workflow-specific success metrics.
• Create a deployment + configuration pattern so new agent workflows can be shipped quickly and maintained over time.
• Establish engineering standards: architecture docs, code review norms, CI/CD, secrets management, and runbooks.

Collaboration

• Work closely with the founders on product strategy + technical direction.
• Partner with delivery leads to translate client constraints into scalable platform primitives.
• Collaborate with other engineers and contractors, and help raise engineering standards as the function grows.

What we’re looking for

• Deep Python experience building and operating production systems (APIs, services, data pipelines) with strong code quality standards.
• Multi-agent systems experience is required: you've built stateful, tool-using agentic applications. We currently use LangChain + LangGraph and LangSmith for tracing, debugging, evaluation, and monitoring, and expect you to work in this stack or ramp quickly.
• Tooling + integrations for agents: strong experience building/operating tool interfaces for agents (API clients, permissions, audit logs), including familiarity with MCP (Model Context Protocol) or equivalent patterns.
• Strong exposure to, or genuine interest in, Marketing & Media workflows (campaign production, compliance, brand governance, media, measurement, MarTech ecosystems).
• Senior backend / platform engineering experience shipping production systems.
• Strong systems thinking: reliability, security, and operational ownership.
• Experience building integrations against third-party APIs and enterprise systems.
• Comfort working in ambiguous, high-ownership environments (0→1, early-team cadence).
• Experience with cloud infrastructure (AWS preferred), IAM/security patterns, and deployment automation.
• Familiarity with enterprise security requirements: audit trails, data residency, SOC2-adjacent considerations, and deploying inside client-controlled infrastructure.

Success looks like (first 90 days)

• Meaningful contributions to the platform architecture and roadmap, aligned to Defyner’s near-term deployments.
• A working “platform baseline” (logging/monitoring + integration scaffolding + deployment pattern) used by at least one live client workflow.
• Evaluation + regression checks in place for core workflows.

Strong signals (any combination)

• You've shipped 0→1 reliability infrastructure (observability, evals, incident response) in production AI systems where errors carry real downstream risk.
• You've built integrations against enterprise systems (CRM, CMS, ad platforms, data warehouses) with proper permissioning and audit trails.
• You've navigated enterprise security procurement, data residency requirements, or compliance reviews in a deployment context.
• You've worked with Marketing/Media brands or platforms and understand the operational rhythms, tool ecosystems, and governance pressures they face.

Compensation + equity

• Compensation: £90–140k depending on seniority, location, and scope, plus performance upside. Final package agreed at offer.
• Equity: 0.75–1.5% pre-Series A, on a standard 4-year vest with 1-year cliff, which is meaningful ownership to reflect real impact.

Benefits

• Remote-friendly (UK/Europe/US) with flexibility around hours and location.
• High-trust environment with significant autonomy and direct founder access.
• Top-of-market learning curve: production agent systems, evaluation/observability, enterprise integrations, and real brand + compliance constraints.
• Clients own the systems we build, and our moat is methodology, expertise, and the platform layer. You're building real infrastructure that clients operate rather than a black box.
• Opportunity to shape the platform, standards, and engineering culture from day one.

Hiring process

• Intro call (fit + motivation)
• Technical deep dive (systems, reliability, integrations)
• Practical exercise or architecture discussion (agent platform scenario)
• Final conversation with founders + references

Logistics

• Location: UK, Europe, or US (remote-friendly; overlap with UK/EU time zones preferred).
• Seniority: Senior IC joining early in the platform engineering function, with room to grow into a lead role.
• Start: Targeting July/August 2026.