Head of Agent Architecture
Pubblicato il 20-08-2026 - Jobgether in Italia
This position is listed on behalf of a partner company, who manages all applications and next steps. Our partner is looking for a Head of Agent Architecture based in Italy.
This is a senior, player-coach leadership role shaping how AI agents are designed, evaluated, and deployed across a next-generation creative technology platform.
You’ll define the technical vision for agent architecture, including multi-agent workflows, task decomposition, state management, tool use, and evaluation systems.
Alongside setting strategy, you’ll stay hands-on with the most challenging architectural problems and establish standards that allow the function to scale.
You’ll build and lead a small team of Agent Architects, developing both their technical capabilities and their ability to solve complex client problems.
The role also has a strong client-facing dimension, with responsibility for strategic discovery and translating ambiguous business and creative challenges into robust technical solutions.
You’ll work across AI Engineering, Product, Sales, and client leadership to balance quality,
reliability, cost, latency, and scalability.
This is an opportunity to help define the infrastructure, practices, and people shaping how AI-powered advertising is created at scale.
Accountabilities
- Define the technical vision and roadmap for agent architecture, covering multi-agent workflows, task decomposition, state management, tool calling, retrieval-augmented generation, API integrations, and related patterns.
- Establish architectural standards and make strategic decisions about what should become reusable platform capabilities versus bespoke solutions for individual clients.
- Design and maintain reusable agents, prompt libraries, orchestration patterns, and frameworks capable of supporting thousands of brands with distinct requirements and voices.
- Lead the systematic evaluation and optimization of agent performance, establishing benchmarking, experimentation, and evaluation practices t