UK remote
Agentic Platform Engineer
About this role
We’re looking for an Agentic Platform Engineer to join our tech team and help build the future of neuro-contextual advertising at global scale. Who We Are At Seedtag, our mission is to transform advertising by proving that effectiveness and user privacy can truly coexist. As the leading Neuro-Contextual Advertising Company, we combine Artificial Intelligence, Natural Language Processing, Computer Vision, and neuroscience to understand not only what content is about, but how it makes people feel and what they intend to do next.
Our proprietary AI, Liz, enables brands to connect with audiences across the open web and Connected TV without cookies or user tracking. Founded in 2014 by two ex-Googlers, Seedtag has grown to 500+ Seedtaggers in 17 countries, backed by €250M in funding, and operates today as a global ad-tech leader. If you enjoy solving complex engineering challenges and building AI-driven systems at scale, you’ll feel right at home here.
Your Challenge As our Agentic Platform Engineer, you will help build and scale Liz Agent, Seedtag's AI agent platform, ensuring reliability, security, and seamless integration across teams. Your main responsibilities will include: Agent Platform Development: Build and maintain the core Liz Agent runtime, including agent execution, context management, and event streaming. Tools & Integrations: Develop MCP tool servers, APIs, and integrations that enable secure access to internal data and services.
Agentic AI Capabilities: Improve agent functionality through memory, multi-model routing, durable execution, and LLM frameworks. Reliability & Observability: Monitor platform performance, troubleshoot production issues, and improve tracing and evaluation capabilities. Platform Engineering: Contribute to APIs, service templates, CI/CD pipelines, and cloud-native deployments that enable teams to build agents independently.
Security & Quality: Implement security best practices, automated evaluations, and regression testing to ensure reliable agent behaviour. End-to-End Ownership: Take responsibility for features from development to deployment and monitoring, collaborating with Data, ML, Platform, and Product teams. Our Core Values Outcome over Output We measure success by impact and value, not by volume of features or lines of code. Failure Is Allowed, Learning Is a Must Experimentation is key to innovation.