Want to know if this job is worth applying to?
Berlin, Deutschland
Onsite
Langdock exists to change the way the world works, bridging the gap between what technology can do and what people actually do with it. We bring all leading AI models into one secure, model-agnostic platform and make them usable across entire organizations. Over 10,000 companies use our platform every day, from fast-growing startups to some of Europe's largest enterprises. Their employees open Langdock to draft strategies, analyze documents, or automate workflows - helping them to work smarter, think more creatively, and reach their full potential.
This is a full-time, 12-month program at our Berlin HQ for people who want to understand AI rather than just use it.
It begins where our product meets the real world: customer support. Many support tickets are a small puzzle about how large language models and cloud software behave in production. Why did the same prompt give two different answers? Why did a long document get cut off? Why is one model fast today and slow tomorrow? Why did an agent call the wrong tool? Why did an API integration time out? You will be the person who works out what is actually going on and explains it clearly.
For most of your first year, the support queue is your home base and your classroom. Few jobs expose you to as many real AI systems, real enterprise environments, and real failure modes, every single day.
The purpose of the year: by month twelve, you should understand our platform and operations deeply, from the chat interface down to the models and infrastructure behind it. A Langdock business team should want to hire you permanently, and you should have found the work that energizes you most.
Customers ask the questions that the rest of the world is still figuring out. To answer them, you will build real intuition for:
How LLMs actually behave: tokens and context windows, sampling and non-determinism, hallucinations, prompt sensitivity, system prompts, and why models from different providers react differently to the same input.
Agents, tools, and retrieval: how agents plan and call tools, how knowledge bases and document search feed context to a model, and where these chains tend to break.
Cloud software in production: APIs, authentication and SSO, rate limits, latency, integrations, multi-tenant enterprise setups, data residency, and access control.
Inference and compute: what happens when a request hits a model, why throughput, latency, and cost trade off against each other, and how running our own inference changes what we can offer customers.
Daily operational tasks: resolving billing issues, reporting bugs and being the point of for our customers.
You do not need to know all of this on day one. You need to be the kind of person who cannot stop asking why until you do.
Onboarding: Learn the platform by supporting it. You go through an intensive onboarding into our products, the underlying models, and the support queue. You shadow your first tickets, get to know our systems, and start resolving real customer issues. Expect plenty of experimenting, plenty of questions, and a steep learning curve.
Post-Onboarding: Own the queue and take on sidequests. Alongside the queue, you start picking up work from other teams: tasks below the threshold for a specialist to own but too valuable for your learning to ignore. This includes taking a sales call for a new lead, running a short customer workshop, building customer use cases, writing internal documentation, building internal agents, and further building out our own AI support tooling. As our inference and cloud offering grows, you will increasingly help customers understand and adopt it.
Front-line customer support. Spend most of your day in the queue, handling customer tickets end-to-end across model behavior, API integrations, enterprise compliance, billing, and access control.
Investigate what is happening under the hood. Reproduce issues, compare model behavior, read logs and API responses, and separate "the model did something unexpected" from "the system did something wrong." Then give users clear, empathetic, and accurate answers.
Explain AI to the people who use it. Help customers understand non-determinism, model limits, context handling, latency, and model choice in plain language, so they can use AI with confidence.
Take on cross-team sidequests. Pick up the real work that falls between teams, like a small inbound sales call, a customer workshop, a piece of documentation, or an internal automation. These tasks are small in scope, but you fully own them.
Escalate to engineering and product. Flag recurring user friction, document edge cases with clear reproduction steps, and report bugs directly to the engineers building our platform and infrastructure.
Improve support with AI. Propose practical improvements to our support workflow and build and test the internal AI tooling that helps resolve routine questions faster.
You start on a 12-month fixed-term contract. The program is designed so that strong performance turns into a permanent contract in a business department that suits your strengths: AI Adoption Management, Sales, Partnerships, Marketing, Recruiting, RevOps, Operations, Solutions Engineering, or elsewhere in the Business team.
We are expanding fast in two directions: across the work people do, with chat, agents, meetings, code, and apps; and down the stack, into our own cloud, compute, and inference. Wherever you land, you will bring something rare: a deep, hands-on understanding of how AI and AI infrastructure actually behave for real customers.
We make the decision together based on your performance in the queue, how you handled sidequests, and whether the team you want to join has a role for you. So far, AI Associates have graduated into Account/AI Adoption Management, Sales and Solutions Engineering. We are looking to broaden the program as a gateway into all business functions.
You are fascinated by how AI works. You have gone down rabbit holes on why models behave the way they do, and you want a job where that curiosity is the core of the work.
You want to understand the stack, not just the surface. Cloud software, APIs, and inference interest you, and you are excited to learn how models are served, scaled, and made fast and reliable.
You are genuinely excited about front-line AI support. You take pride in helping customers, write with patience and clarity, and enjoy working through a busy ticket queue.
You think in systems and love analysis. You enjoy breaking down tricky technical or operational problems and figuring out what is really happening under the hood.
You show high agency and full ownership. You see a ticket through from first message to resolution and dig into root causes so the same issues do not keep coming back.
You experiment constantly. You try new models, automations, and tools, and you instinctively look for ways technology can make work easier.
You communicate clearly and professionally in English. German is a plus and useful in the queue.
You are kind and care about the people around you.
Availability to start full-time and work 100% on-site at our Berlin office for the full 12 months.
A work permit for Germany.
Technical or analytical grounding. Studies or hands-on experience in Computer Science, Data Engineering, or a related field, or a track record of building technical projects, is a plus but not a requirement. Curiosity about how AI and cloud systems work is.
You will work from our Berlin office at Greifswalder Straße 212, where all of us work together in person, surrounded by people building AI products and infrastructure every day. Lunch and dinner are catered in the office. We look after ourselves, and gym membership and coaching resources are part of the standard package. We value Calm Urgency, Ambitious Execution, and Caring Ownership. The vibe is calm but intense.
Salaries are transparent and tied to levels, not negotiation. We narrow down the expected salary range early in the process.
We move fast, and most hiring processes finish within two weeks. If you want to spend a year truly understanding AI, from the conversation down to the compute, we would love to hear from you.