CONTACT
CONTACT

AI Automation Architect

Design and build the AI workflows and agents that do real work inside our clients’ SaaS products, and inside our own delivery.

What you’ll do:

  • Map client workflows with designers and product leads, and decide what to automate, what to assist and what should stay with people.
  • Design agentic workflows: the trigger, the steps, the tools the agent can use, where a human approves, and what happens when something fails.
  • Build working prototypes and internal automations with LLM APIs, orchestration tools (n8n, Make, LangGraph or similar) and the Model Context Protocol (MCP).
  • Define evaluation sets, guardrails and quality thresholds before anything ships, and measure accuracy, cost and latency against them.
  • Write technical design documents and decision records that client engineering teams can build from without guesswork.
  • Work with experience designers so every automation has a clear interface: visible status, explainable outputs, easy correction.
  • Build and maintain Hue & Machine’s internal automation stack: research synthesis, design-to-documentation, QA checks, client reporting.
  • Advise clients on AI feasibility, risk, privacy and cost, translating trade-offs for founders and engineers alike.

What we’d like to see:

  • 4+ years in automation, solutions engineering, technical product management or full-stack development, including 2+ years hands-on with LLMs.
  • Shipped AI workflows or agents used by real people, with the results to show for it.
  • Strong grasp of LLM behavior: prompting, context management, tool use, structured outputs and their failure modes.
  • Practical experience with APIs, webhooks and integrations across SaaS tools (CRM, support desks, product analytics, data warehouses).
  • Scripting in Python or TypeScript.
  • Systems thinking: you see the inputs, logic, edge cases and costs behind a flow before you build it.
  • A clear opinion on when AI should not be used, and the confidence to say so to a client.
  • Can explain complex technical concepts to designers, executives and engineers.
  • Business-level English.

Nice to have, but not required:

  • RAG, vector databases or multi-agent orchestration.
  • A background in product or service design, or close work with design teams.
  • Knowledge of AI governance, GDPR and EU AI Act requirements.
  • Evaluation pipelines or LLM observability (Langfuse, LangSmith or similar).

Why working with us will be a great fit:

  • Smart projects first. Work that sparks curiosity and rewards clear thinking, from concept to polished execution.
  • Paid AI tools. Use what helps you do your best work. We cover it.
  • No bureaucracy. Clear briefs, structured support, fast feedback. Meetings only when they matter.
  • Global clients, close team. Clients across the US and Europe, and a tight-knit team that values autonomy, clarity and good humor.
  • Room to grow. Regular design reviews, 1-on-1 mentorship and internal knowledge sharing.

Recruiting process:

  1. Recruiter screen
  2. Technical interview
  3. Offer

For each example, show the problem, how you designed the workflow, how you evaluated it and what it changed.

Apply for this job

Send your CV in English and examples of AI workflows or agents you’ve built (a demo, a repository, a case write-up or a short video) to . We can’t review applications without at least one example. If there’s a fit, we’ll be in touch.