Automation Engineering
AI Automation Engineer
Design, build, and operate the AI automations and agents that do measurable work — across client engagements and our own operations.
Role at a glance
| Reports to | TBD |
| Location | Remote — anywhere in Europe (quarterly meetups, occasional client travel) |
| Level | Junior – Mid-Level |
| Function | Engineering · Automation & AI |
About the Role
PromptPartner is looking for a Junior to Mid-Level AI Automation Engineer to design, build, and operate AI-driven automations and agents across our client engagements and internal operations. You will work hands-on with the Claude Desktop application and the Claude API, local and hosted LLMs, and Model Context Protocol (MCP) integrations to turn manual, repetitive workflows into reliable, monitored agentic systems.
This is a builder role: you will own solutions end to end, from prototyping an agent to hardening it for production use — across delivery, operations, finance, and everything in between. And when your work ships into a client environment, it does measurable work their CFO can point to.
To be very honest, we’re not for everyone. We move fast. We make decisions quickly. We value output over hours. If you thrive on ownership and want your work to matter — welcome home.
What You’ll Do
- Design, build, and deploy AI automations and agents using Claude Desktop and the Claude API for real business workflows.
- Partner with delivery, operations, finance, and GTM stakeholders to identify automation opportunities and translate them into agent-based solutions.
- Develop and maintain MCP servers and/or an MCP gateway to connect agents securely to internal and third-party systems — CRM, ERP, data warehouses, ticketing, and client environments.
- Stand up and manage local LLM environments and Retrieval-Augmented Generation (RAG) pipelines for use cases requiring data privacy, low latency, or offline operation.
- Build and maintain an agent harness (orchestration, tool-calling, memory, guardrails, logging, evaluation) around local and hosted models.
- Write production-quality Python (and other languages as needed) to support agent tooling, data pipelines, and integrations.
- Establish testing, monitoring, and evaluation practices for agent reliability, cost, latency, and output quality.
- Document architecture, runbooks, and guardrails for all deployed automations and agents.
- Stay current on the AI/agent tooling landscape and recommend improvements to the automation stack.
Required Skills & Experience
- Automation & agent development with Claude. Hands-on experience building automations and agents with the Claude Desktop application and the Claude API.
- Domain experience. Prior work building agents or automations for real business use cases — operations, finance, GTM, customer support, or internal tooling.
- Programming. Proficient in Python, plus familiarity with other languages relevant to AI/automation work (e.g., TypeScript/JavaScript, Bash).
- MCP (Model Context Protocol). Experience building and managing MCP servers directly, or operating agents through an MCP gateway.
- Local LLMs & RAG. Practical experience deploying and working with local/self-hosted LLMs and building RAG pipelines — chunking, embeddings, vector stores, retrieval evaluation.
- Agent harness design. Proficient in building an agent harness around local LLMs: orchestration, tool use, memory, guardrails, and evaluation loops.
- Remote anywhere in Europe. Comfortable with occasional client travel.
Preferred Qualifications
- Experience with vector databases (e.g., Pinecone, Weaviate, Chroma, pgvector)
- Familiarity with enterprise systems such as CRM, ERP, billing, or support tooling
- Experience with CI/CD, containerisation (Docker), and cloud platforms (AWS/Azure/GCP)
- Understanding of AI safety, data privacy, and security best practices for enterprise agent deployments — including EU AI Act considerations
Why PromptPartner
- Build for production. What you ship runs inside real client environments: European software companies, PE funds, consultancies and professional services. It does measurable work their CFO can point to.
- Real ownership. You own solutions end to end, from prototype to hardening, and you see what they do for the client.
- High-bar team. Everyone here is someone you’ll want to learn from — operators, engineers, and strategists at the top of their craft. No politics. No ego wars.
- Direct access. Operators and executives, with none of the layers, status meetings, or process overhead of a traditional firm.
- Small, early team. Your judgment shapes the product, the playbooks, and how PromptPartner scales.
- Early to the thesis. We bet on this category before it was obvious — and we’re already putting a euro figure on AI for companies doing nine figures of revenue.
What You Get
- Remote-first — work from anywhere in Europe
- Flexible hours — output over hours. Limited meetings, deep-work focus.
- Learning — annual learning budget and mentorship from day one.
- Equity — shared success. Your work is in the company.
- Gatherings — quarterly in-person meetups with the whole team.
- Real impact — your work powers AI systems for real businesses.
How to Apply
Send your CV and a short note on the most interesting agent or automation you’ve built — what it does, where it runs, what it saves — via promptpartner.ai/careers.
Apply now · promptpartner.ai/careers
We are an equal opportunity employer. All qualified applicants will receive consideration for employment without regard to race, colour, religion, sex, sexual orientation, gender identity, national origin, disability, or protected veteran status. Employment is subject to a standard confidentiality and non-disclosure agreement.
PromptPartner AI — AI that does measurable work inside your company.
