Engineers who don't wait for a spec (Forward Deployed Engineer)

It's Monday, your first week with a new client.

The finance team of a US company closes the month in spreadsheets and in a system nobody has documentation for anymore. The controller shows you how they do it. For half of the steps, nobody can tell you why they are done this way. There is no spec. The first prototype on their real data is due in a few days.

Most developers would ask for a specification.

You first figure out what is supposed to happen, and only then teach it to a machine. And when you realise that part of it doesn't need to be built at all, you say so.

That is what a Forward Deployed Engineer is: the engineer who is the face of the project at the client, not someone who receives tickets. You lead discovery, design the architecture, write the code, put it in the hands of real users and tune it until it is genuinely useful. Investors at a16z called this role "the hottest job in tech".

What you will work on

• AI automation in companies. Month-end close, compliance checks, data migration and reconciliation. Agents and validation tools do the work and people only handle the exceptions.

• Rebuilding legacy systems. You rebuild a critical, often undocumented system on a modern stack. AI speeds up reverse-engineering, code translation and test generation.

• AI in the client's engineering team. You sit in their team and set up tools, agents and evals so that the team knows more after you leave and doesn't depend on you.

• New AI products from scratch. From discovery to production.

How Sudolabs works

• One client at a time. A small team of 3–5 people: a product lead, AI and data engineers, a designer.

• Remote-first. No mandatory office time. There is an office in Košice if you want to come in.

• Fast prototypes. Solutions reach users quickly and are tuned on their data.

What we won't promise you

• We won't promise you a finished spec. Requirements change and you will have to find out part of them yourself.

• We won't promise you clean code or documentation. Client systems are old and tangled, which is why they call us.

• We won't promise you'll only be coding. You'll spend a lot of time with the client's people, from controllers to the CTO.

• We won't promise calls only in our time zone. Most clients are in the US.

• We won't promise zero travel. Sometimes it pays to spend the first week or two on site with the client. It is the exception, not the rule, and it is planned with you.

What we do promise: you will see your work in production and see how it changes the way a business runs. In weeks, not quarters.

This might be you if:

• you can design and ship a production system, not just a prototype, and you are comfortable across the full stack (TypeScript / Node / React and Python),

• you really work with AI: LLMs from idea to production, agents, context engineering, RAG and evals that tell you whether the output is good enough,

• you can see where the leverage is in a messy process and say what should not be built,

• you can run a working session and explain a technical trade-off to someone who isn't technical,

• ambiguity doesn't stop you and you own the result from start to finish,

• you remember that you work with the client's money and you keep the scope under control,

• you speak English fluently and working with clients in other time zones doesn't bother you.

Nice to have, not required: experience in a client-facing engineering role (forward-deployed, solutions, applied AI), knowledge of finance, insurance, data migration or developer tooling, data engineering (SQL, ETL, warehouses), ML beyond off-the-shelf LLMs, AWS and serverless.

What you really don't need: the title "Forward Deployed Engineer" on your CV. What matters is that you have already built production systems that real people use.

What you get:

• A front-row seat to AI in practice. Real production systems and a new hard problem every few months.

• EUR 25–46/hour depending on experience, more for exceptional people. Cooperation on a trade license (invoice).

• 31 paid days off and 20 paid days a year for learning, during working hours.

• Freedom. Flexible hours, work from anywhere.

• Equipment. A MacBook or whatever you need.

• Pet-friendly office in Košice city centre.

• Free parking at our HQ.

• Sudo apartment available for those living outside Košice.

• People. Internal tech talks, team buildings and company retreats.

📖 Not sure what an FDE actually does? Read our article (in Slovak): Forward Deployed Engineer: najžiadanejšia práca v AI a ako sa ňou stať