AI Referent · Automation & Operations Engineer · AI agents · n8n · Integration & training
I put AI to work, with method: I design agents and automations, integrate AI tools into business workflows, and train teams to use them. I've already designed and led an internal AI masterclass (7 modules) and built an agent-orchestration hub to automate content publishing.
I start from the business need, identify the productivity gain, then ship the automation or agent that delivers it, and stay accountable for keeping it running. I strictly separate orchestration (n8n) from business logic (a decoupled API), for automations that are reliable, testable and reusable, not throwaway hacks. Priority on framing, documentation and training the teams.
An automation hub I designed to automate content creation & social publishing: n8n orchestrates the “when”, a decoupled Fastify API keeps the “how”, Postiz publishes. Built to integrate models, tools and agent workflows.
A full program for a disciplined use of AI in a company: from the LLM mental model to context engineering, from prompting to agent workflow, through quality & security. With the templates that make output predictable.
Integration of latest-generation coding agents into dev workflows: framing, context engineering, reviews, and use on legacy code, for real, controlled productivity gains.
End-to-end scripted deployment (pre-checks, DNS, TLS, validation, maintenance) and self-hosted internal services (Dolibarr ERP, project management, mail): I optimize internal operations, not just the code.
A complete platform I architected and deployed (NestJS, Expo, Next.js, PostgreSQL). Proof that I master the full technical chain any serious automation rests on.
I test new AI technologies as they ship (models, tools, agents), assess their real value, then pass them on to the team. Analytical mindset, focused on concrete gains over hype.
AI only has value if it saves teams real time and holds in production. I aim for automations that are reliable, documented and adopted, never demos that break at the first edge case. Curious, I keep upskilling on agents, RAG and AI FinOps.