Hi — we're Hypefy, a Croatian-based start-up.

We're building AI agents that run influencer marketing end-to-end: campaign creation, creator selection, outreach, content checks, payments, reporting — all automated, all at scale.

AI Enablement is Hypefy's in-house AI team. A small, hands-on group that works across every other department - Sales, Marketing, Customer Success, Campaigns, Product - to figure out where people's week gets stuck, and then actually build the agents and tools that unstick it. 

The job covers the whole arc: understanding how a team really works day to day, figuring out what's actually slowing them down, building the smallest thing that fixes it, and then proving it worked using that team's own numbers.  

If you are a mid-to-senior engineer and you would like to take something from a recurring pain point to a two-minute tool, with the data to back up that it helped, this is for you. 

What you'll work on

• Own initiatives end to end: sit with a department, map how the work actually flows, find the constraint, ship the smallest tool or agent that moves it, and measure that it did — on the department's numbers, not just hours saved

• Build the agents people use every day — Claude Code skills and plugins, MCP servers, workflow agents, small internal apps

• Design and run the harnesses those agents live in — tools, context, guardrails, evals, tracing — so a sales rep or a campaign manager can trust the output without reading the code

• Engineer the context, not just the prompt: give agents the right company knowledge, data access and memory, and keep it current as the company changes

• Drive adoption: demo it, onboard people, take the feedback, and change the process around the tool — a tool nobody uses moves no number

What We're Looking For

• You build with AI and have shipped something real with it — an agent, a tool, an automation people actually use

• Fluency in the Anthropic and/or OpenAI ecosystem: agentic coding tools such as Claude Code or Codex, the APIs, tool use, MCP. You know what a skill, a plugin and a harness are because you've built or configured them

• Context engineering as a habit: what an agent sees decides what it does, so you design its context — documents, data access, memory, tool descriptions — on purpose

• Engineering fundamentals to back it up: production-quality TypeScript or Python, APIs and data, git, and the ability to debug a system you didn't write. A classic engineering background is a big plus; what matters is that what you build holds up in production

• A problem-solver first: you start from the problem and the number, not from the tool, and you can tell when the answer is an agent and when it's a rule change or a spreadsheet

• Ownership: you take a problem from "here's the pain" to "here's the result", decide without being pushed, say early when something is off, and show your work — in person or through your agents — so people outside engineering get what you built and why it matters

Nice to Have

• Designing or extending agent harnesses: custom runtimes, subagent orchestration, workflow engines, hooks, evaluation pipelines

• Observability and evals for LLM systems (Langfuse or similar), and the discipline of measuring output quality, not just whether it ran

• Data and integration work: SQL over a warehouse (we use DuckDB and MotherDuck), CRM and marketing tooling (HubSpot), Microsoft 365, Azure or Vercel

• You've run something of your own — a product, a company, a side business — or otherwise have an entrepreneurial streak

• Background in SaaS, marketplaces, go-to-market tooling, or influencer and creator platforms

Why Join Hypefy

You'd be shaping how a fast-growing company uses AI from the inside, with a mandate from leadership and a direct line to every department head. The work is judged by the numbers it moves — demos booked, deals unstuck, campaigns delivered — not by how sophisticated the stack is, and you'll see the effect of what you ship within weeks, with the people who use it a few desks away. If you want to work at the edge of what agentic tooling can do, and grow a company's capacity ahead of its headcount, this is that kind of role. 

You'll have flexible hours, you'll be in our Zagreb office most of the week with the team, and you'll be doing it alongside people who are ambitious, low-ego, and genuinely enjoy the work.