AI automation jobs: what an AI ops specialist actually does

AI automation jobs are roles where you build and run AI workflows and agents for your own company's teams: sales, operations, finance, people, support. The job is real, but new and still rare. On Tokn, as of 27 October 2026, we found 19 live postings for it at 14 companies, 11 in Europe and 8 in North America, under more than a dozen titles.

Our definition: a posting counts when its title and text say the job is to build AI automations for the company's own teams, or to lead their adoption. Customer-facing builders don't count, and neither does RPA, factory or test automation. For the broader picture, start with the 8 paths into AI.

In short

• An AI ops specialist finds repetitive work inside the company, redesigns it, builds the agent or automation, then monitors it and trains people.
• Tokn found 19 live postings at 14 companies as of 27 October 2026 (11 Europe, 8 North America). Most are lead roles.
• Ads ask for agents, Python and integrations before any no-code tool. n8n is the platform with real pull, all of it in Europe.
• In North America, the 8 posted base bands run from $134k to $405k. In Europe, one posting showed pay: €50k–€80k at Dust in Paris.

What does an AI ops specialist do?

An AI ops specialist turns manual work inside a company into AI workflows that run reliably, then keeps them running. The users are colleagues, not customers.

The core loop

First, find the bottleneck. Anthropic's AI Operations Engineer posting wanted someone who can "look at a messy manual workflow, find the actual bottleneck, and redesign it rather than just automating it as-is". Then build fast, with "a working v1 in front of them in days, not quarters", using "Claude, MCP connectors, and API integrations".

Then run it. The same ad asked for "reliable, monitored automations" with "guardrails and human-in-the-loop checkpoints". Supabase's Head of AI Native Operations posting went further: "Replace workflows, don't assist them."

Who they build for

Mostly go-to-market and operations teams. Dust's Revenue AI Operations role was about building "AI-powered workflows for the entire GTM engine". ElevenLabs described its internal role as "acting as a forward-deployed engineer across our GTM, Operations, and Finance teams", wiring AI into Salesforce, Slack and Ashby.

Some roles sit closer to IT. Legora's Automation Engineer in New York covered "onboarding and offboarding, access provisioning, SaaS administration". LangChain's Agent Systems engineer ran an internal GTM agent, including "the weekly eval suite" and cost and latency tracking by model.

The part nobody advertises

A good share of the job is control. Of the 19 postings, 8 mention security, permissions or guardrails, 4 human-in-the-loop checkpoints and 4 evals. Supabase listed "least-privilege credentials, hard policy limits, escalation thresholds, audit logs". An agent that can edit the CRM or grant access needs someone who owns what it may do.

The AI automation job titles companies use

There is no standard title yet. A sample from the 19 postings:

Many are "Lead", "Head of" or "Owner" titles because companies are creating the function for the first time. When you search, try "AI ops", "AI enablement", "AI transformation" and "automation engineer", not only "AI automation".

Not the same job

Forward deployed engineers and deployment strategists build AI for customers. AI engineers build the product. The AI ops specialist builds for colleagues.

This matters when you browse Tokn. Our AI ops and automation jobs category is mostly customer-facing deployment work: forward deployed engineering, solutions, deployment strategy and support. The internal automation job is a small part of it, and several of the 19 postings sit in other categories.

Who is hiring for AI automation jobs?

AI-native companies and automation vendors hire first, and they practise on themselves.

• Europe (11 postings, 10 companies): n8n (two remote roles from Berlin), Dust, Brevo and Equativ in Paris, Doctolib in Berlin, Supabase and ElevenLabs (remote), Harvey in Dublin, Legora in London, GitLab (remote UK).
• North America (8 postings, 6 companies): Anthropic, Scale AI and LangChain in San Francisco, Legora in New York, MongoDB in Palo Alto, GitLab (remote US).

Vendors have a reason to show their tools at work. Dust markets the people who build agents for their teams as "AI Operators", Dust's own term rather than an industry title. n8n, which raised a $180m Series C at a $2.5bn valuation in October 2025, is hiring to spread AI use inside its own teams.

The jobs below come from Tokn's wider AI ops and automation category, so expect deployment roles next to the internal ones.

What AI automation job ads ask for

Agent-building, Python and integrations come before any no-code platform. Counts across the 19 postings (minimums, since we store shortened text):

n8n is the one platform with real pull, and all 6 mentions are European. Zapier showed up in 2 postings. Make appeared in no active posting on Tokn, though we mostly track AI-native tech companies; smaller businesses and agencies that use it are outside our coverage. For skills across all roles, see the most in-demand AI skills, and for agent work in general, agentic AI jobs.

Code or no-code?

You don't always need to be a software engineer, but you need to be comfortable near code. Python appeared in 9 of 19 postings. Even Legora's code-light Automation Engineer ad said "you can read generated code when troubleshooting calls for it", and the Equativ internship asked for Python alongside n8n. If you come from a non-coding role, see AI jobs without a technical background.

Seniority

Entry points are scarce. Of the 19 postings, 6 were lead roles, 2 senior, 1 mid and 1 junior (an Equativ internship in Paris starting February 2027); 9 didn't say. Experience was phrased as proof, not years: "proven experience building and shipping automations… in production" (ElevenLabs).

How much do AI automation jobs pay?

In North America, all 8 postings showed a band, from $134k to $405k base. In Europe, only 1 of 11 did.

That is n=8, all at well-funded AI companies, so read it as a list, not a market rate. These are posted base bands, excluding equity. Engineering-heavy roles sit at the top, ops-leaning ones lower.

In Europe, the single posted band was Dust's Revenue AI Operations role in Paris: €50k–€80k. One data point says nothing about the market; for the wider picture, see AI salaries in Europe.

How to get into AI automation

Most people come in sideways, from a job where they already automate things. The ads point to five doors:

1. RevOps and GTM systems. A Cognition GTM Systems Engineer posting, live as of 27 October, asked for Salesforce, SQL, Zapier and "agent-based workflows". GTM engineer itself is a young, contested title.
2. Support ops and IT. OpenAI's IT Systems Engineer posting covered "internal tools, automation, and autonomous workflows".
3. Customer-facing deployment. ElevenLabs framed its internal role as forward deployed work pointed inward.
4. No-code automation. n8n's enablement role centred on workshops and build-alongs for non-technical teams.
5. Software engineering. Scale AI's and Anthropic's AgentOps roles are engineering jobs with internal users.

A portfolio that works

The ads reward things that run in production, so show that:

• 2 or 3 automations that real people use, with before and after numbers.
• Error handling, and a human checkpoint where a mistake would be costly.
• A short write-up of what you chose not to automate, and why.

To build agents hands-on, getting started with Claude Code is a practical first step.

Where the role is heading

Companies are raising the AI bar for everyone, which creates work for the people who set the tooling up. Zapier says it assesses the AI fluency of every new hire, and its hiring page allows AI in applications. Back in April 2025, Shopify's CEO made "reflexive AI usage" a baseline expectation, as The Logic reported. We think the dedicated role will spread, but most companies still fold this work into existing ops jobs.

Find AI automation jobs on Tokn

Browse AI ops and automation jobs knowing most are customer-facing, and search for "AI ops", "enablement" and "automation engineer". Because the titles vary so much, the simplest way not to miss the rare internal openings is to set a job alert.

FAQ

Do you need to code for AI automation jobs?

Not always, but it helps. Python appeared in 9 of the 19 postings Tokn found as of 27 October 2026, and APIs or integrations in 8. Even code-light roles expect you to read generated code.

Is n8n, Zapier or Make enough to get hired?

A platform helps, but it is rarely what the ad asks for first. n8n appeared in 6 postings, all in Europe, Zapier in 2 and Make in none. Agents, integrations and judgement about what to automate matter more.

How much do AI automation jobs pay?

In North America, the 8 postings with a band ranged from $134k to $405k base, all at AI companies. In Europe, Tokn found one posted band: €50k–€80k at Dust in Paris.

Are there entry-level AI automation jobs?

Very few. Of the 19 postings, one was an internship and most were lead roles. The usual way in is from RevOps, support ops, IT or engineering, with automations you have already shipped.

Are AI automation jobs remote?

Some are. 6 of the 19 postings were fully remote, at n8n, Supabase, ElevenLabs and GitLab, while 5 were explicitly onsite.