Hebno is looking for an intern to turn our outbound plan into working systems: email sequences, CRM updates, LinkedIn automation and AI agents that run without someone pushing buttons.

About Hebno

Hebno is the only AI a consultancy needs: one platform for research, analysis, drafts and decks, working from the firm's own knowledge and approved sources.

Most consultancies already use AI, but in pieces: a chat tool here, a research tool there, knowledge scattered across drives and past decks, and no clear way to check what the AI got right before it reaches a client. Hebno brings that into one place, so teams get to a reliable first draft faster and can review the evidence behind it before delivery.

What the product covers

• Source-backed by default: research and first drafts from approved sources, with evidence, assumptions and gaps visible for review

• Connected knowledge: the firm's prior work, methods and permitted client material, brought together with external sources

• Reusable workflows and assistants: turn a method that works into a repeatable process, so one good project becomes the standard

• The right model, at a cost you control: the best model for each task within what the firm allows, with usage visible by team and client

Who we sell to

Consultancies that deliver research, analysis and advice to clients: research and insight consultancies first, plus strategy, creative and advisory firms. The buyers are usually founders, managing partners, practice leads and operations leads who want their team using AI well, not just using it.

What this role is about

We design the outbound plan; you make it run. The founders decide who to reach, what to say and when. Your job is to build the systems that carry that out reliably.

"GTM engineering" is a new title and means different things at different companies. Here it means part builder, part operator: you write automations and small agents that connect our tools, so outbound happens on schedule, lands in the CRM correctly and doesn't depend on someone copying and pasting.

A typical outbound flow you would own end to end:

1. Build a target list of consultancies that match our customer profile

2. Enrich each account and contact with the data a good first message needs, such as their practice areas, the kind of client work they deliver and how they talk about AI today

3. Enroll them in the right sequence, with personalisation drafted by AI and checked by rules

4. Send email and LinkedIn touches on schedule, within safe sending limits

5. Log every touch, reply and status change in the CRM automatically

6. Route replies to a founder fast, and stop the sequence when someone answers

7. Report weekly on what was sent, what got replies and what broke

What you'll do

Build the outbound machine: wire up email sequences, LinkedIn touches, CRM updates and lead enrichment so they run on their own. Then add AI agents that research accounts, draft openers and sort replies while you sleep. If it's repetitive, your job is to make a robot do it.

You won't be writing the strategy or the copy alone. You will work from what the founders define, and your feedback on what's working will shape the next version.

What we're looking for

Must have

• Willingness to learn fast and figure things out

• Comfortable using Claude, Claude Code and Claude routines

• Clear written communication and the habit of saying when you're stuck

Nice to have

• Experience with tools like Clay, n8n, Make, Zapier, HubSpot, Apollo or similar

• Hands-on use of LLM APIs or agent frameworks

• Interest in sales, startups or how B2B companies find customers

• Familiarity with consultancies or research and insight work

No specific degree or coding background is required. Curiosity matters more than a CV.

What you get

This internship is unpaid, so we want to be clear about what you do get in return.

• Learning: hands-on GTM engineering, a skill set that sits between sales, ops and AI, and few people get to practise early.

• Unlimited Claude: full access to Claude and Claude Code for your work, no usage caps to worry about.

• An ambitious team: work with one of the most ambitious and fast-growing AI startups in Denmark.

• Real ownership: the systems you build will run our actual outbound. You won't be handed a sandbox project.

• Direct access to leadership: you work with the people making the decisions, and you see why they make them.

• Work you can show: with our approval, you can describe and demo what you built in your portfolio and interviews.

• A reference: if you do good work, we'll say so to future employers, specifically and in detail.

• Flexibility: remote or in the office, with hours you can fit around studies or other commitments.

How you can grow

• First weeks: learn our tools, our customer profile and the outbound plan, and ship your first small automation

• First month or two: own one channel or workflow end to end, such as email sending plus CRM logging

• After that: take on more of the system, including agents and reporting, and suggest improvements based on results

How far you go depends on what you ship. If the role grows into something bigger and you're a strong fit, we'll talk about it openly, but we can't promise a paid position or a job offer today.

What to expect

We're a fast paced team. Structure is light, and you'll often figure things out from docs and trial runs before asking.

The role is still taking shape. We know what needs to get built, but the exact tools and scope may change as we learn what works.

It's not a sales job. You won't be cold calling or carrying a quota. You build the machine; the founders talk to the leads.

Some of it is unglamorous. Cleaning data, debugging a failed sync and checking sending limits are part of the work.

How to apply

Send a short note to our email with:

1. A few lines about you and why this role interests you

2. Anything you've built or automated with Claude, if you have it (a link, Loom or screenshot is fine)

3. Your availability: start date and hours per week

Next steps are a 30-minute call with a founder, then a small take-home task: sketching how you would automate one step of an outbound sequence.