AI Engineering & Automation Specialist (on-site)

Your roleEvery internal team at Cloover, Ops, Energy, Sales, CS, Finance, has work that AI agents can do faster and better.

You'll build the systems that make that real: agentic pipelines, MCP servers, and internal tools that change how we operate day-to-day, not in some future roadmap.

Success in this role means colleagues stop doing the work you've automated, and start doing the work only they can do.

You'll report to Florent (Head of Operations & Energy) and work within a team of amazing people alongside.

Your missionDesign, build, and ship end-to-end agentic systems, RAG pipelines, embeddings, retrieval, prompt engineering, and evaluation, that automate Cloover's highest-volume internal workflows.

Build and deploy custom MCP servers and integrations connecting HubSpot, Slack, Notion, Aircall, and other tools, so data flows where it needs to without manual work in between.

Write production-grade Python and TypeScript with sound schema design, security, and authorization, code that ships and runs, not prototypes that sit.

• Own internal tooling end-to-end: spot the gap, build the tool, and deploy it on Firebase, Vercel, or GCP with proper version control and CI/CD.

Build custom Claude Code.

Habilidades

, plugins, and connectors, then onboard colleagues into AI-native workflows that they actually keep using.Partner with Ops, Energy, Sales, CS, and Finance to identify the highest-leverage automation use cases, then personally implement them.Your profileWe care more

Acerca de

how you think and what you've built than where you've been

Here's what will set you up for success in this role:You've shipped agentic systems, RAG pipelines, or AI-powered tools, in school, side projects, hackathons, or work, not just experimented with prompts in a chat window.

You write production-grade Python and/or TypeScript and can defend the trade-offs in your code.

You're comfortable wiring REST APIs, webhooks, and authentication across third-party tools (HubSpot, Slack, Notion, or similar) and can model.