AI Agent Architect - Remote Conversational Platforms

Tamarind Intelligence; Madrid; Presencial; EUR 90.000 - 130.000; Jornada completa

Ventajas ofrecidas por este puesto de trabajo

Occasional Madrid meetings

Descripción de la vacante

Shakers seeks an AI Agent Architect for a conversational AI platform. This 100% remote, full-time role requires hands-on design of agentic architectures and guardrails, with occasional on-site meetings in Madrid.

You’ll direct Claude Code-enabled builds, shape tone across chat and voice, and monitor performance post-launch. Fluent Spanish and strong English are needed, plus a track record in deploying AI agents and safeguarding against hallucinations.

Formación

• Proven experience designing end-to-end agentic systems and flows.
• Experience preventing hallucinations with clear criteria for responses.
• Product/UX mindset linking business goals to technical solutions.
• Hands-on use of agent coding tools to direct build work.
• Portfolio deploying virtual assistants or AI agents.

Responsabilidades

• Frame agentic architectures for chat and voice channels.
• Define guardrails: user recognition, queue management, human escalation.
• Set boundaries of each agent: when to answer/retrieve/ admit ignorance.
• Translate business goals into concrete rules and flows.
• Shape tone and channel adaptation across all channels.
• Direct coding tools to build and validate agent behavior.
• Monitor post-launch KPIs and iterate on edge cases.
• Guide client decisions and translate them into implementation.
• Coordinate with Product, DevOps, CS and Support.
• Train client teams to manage and optimize AI agents.

Conocimientos

Agentic architecture

Claude Code

Guardrails design

Escalation decisioning

Product collaboration

API & integrations

Privacy & security basics

Reading-level Python

EU GDPR / AI Act

Herramientas

Claude Code

Descripción del empleo

Shakers seeks an AI Agent Architect for a conversational AI platform. This 100% remote, full-time role requires hands-on design of agentic architectures and guardrails, with occasional on-site meetings in Madrid.

You’ll direct Claude Code-enabled builds, shape tone across chat and voice, and monitor performance post-launch. Fluent Spanish and strong English are needed, plus a track record in deploying AI agents and safeguarding against hallucinations.