A Forward Deployed Engineer (FDE) is a software engineer who builds AI systems on site at the customer and takes them into production, accountable for the result in day-to-day operation – not just for a concept or a feature. In Germany, providers such as Anthropic, OpenAI and Databricks now advertise the role in Munich or Berlin, and in October 2026 Anthropic launched the Claude Frontier Academy, a dedicated training programme for this kind of engineer.
This article focuses on Germany: how the role differs from related jobs, what public job postings reveal about demand, which skills are required, how to get into the role, what the Frontier Academy actually is – and what companies can do if they do not want to build their own FDE team. Where the model comes from and why it is being copied everywhere right now, we have already covered in detail: Forward Deployed Engineers: Why the Palantir Model Is Becoming the Operating System of Enterprise AI.
Key points at a glance
- Role: An FDE sits with the customer, works in the customer's systems and with the customer's data, and brings an AI system into operation. The FDE is measured by everyday use, not by handing over a document.
- Demand in Germany (as of 4 November 2026): Anthropic is looking for a Forward Deployed Engineer in Munich who is a native German speaker, and for its first FDE manager for DACH. OpenAI is looking for an FDE in Munich, Databricks for FDEs in Berlin and Munich.
- Salary: The postings by the large providers mentioned here state no salary. We did not find a reliable salary survey for FDEs in Germany.
- Frontier Academy: According to its announcement, Anthropic is investing 100 million US dollars and aims to train 10,000 "Frontier Deployed Engineers" by the end of 2027. Participation is by nomination from one's own organisation only; cohorts are running in San Francisco, New York and London. Germany is not mentioned.
- Without an in-house team: Mid-sized companies can use the way of working with an external partner – narrowly scoped, with a clear handover.
What a Forward Deployed Engineer is – and what it is not
You will find a short definition in our glossary under Forward Deployed Engineer. The job postings describe the role remarkably consistently. In its Munich posting, Anthropic writes that an FDE "embeds directly with our most strategic customers", builds production applications with Claude inside the customer's systems and delivers artifacts such as MCP servers, sub-agents and agent skills that run in production workflows (Anthropic, Forward Deployed Engineer, Munich). For Munich, OpenAI states that FDEs own "discovery, technical scoping, system design, build, and production rollout" and are measured by production adoption and measurable workflow impact (OpenAI, Forward Deployed Engineer – Munich).
In practice, the boundaries with related roles are often blurred. A guide:
| Role | When it works | What it delivers | What it is measured by |
|---|---|---|---|
| Forward Deployed Engineer | After the decision, embedded with the customer | A running system in the customer's environment | Use and impact in operation |
| Solutions Engineer / Applied AI Architect | Before the purchase decision, advisory | Architecture proposal, demo, proof of value | The customer's technical decision |
| Consultant | Before or alongside implementation | Analysis, recommendation, roadmap | Quality and acceptance of the recommendation |
| Software engineer (product) | In their own product team | Features to specification | Function according to specification |
How fluid the boundaries are is shown by Anthropic's Munich posting for an "Applied AI Engineer, Enterprise": this role guides large German companies up to the decision for Claude and then explicitly hands implementation to a systems integrator, "while the partner does the build" (Anthropic, Applied AI Engineer, Enterprise). The FDE, by contrast, builds personally. For assessing offers, this means: what matters is not the title, but whether someone has access to your systems, writes code and is accountable for go-live.
Demand in Germany: what the job postings show
There are no official statistics on FDE positions. On 4 November 2026 we therefore evaluated the public job boards of three large providers directly (Greenhouse and Ashby interfaces, no aggregators):
- Anthropic lists 16 positions based in Munich. Among them a Forward Deployed Engineer (including "Native German speaker with fluent English proficiency", 8+ years of experience, 25–50% travel), a Manager, Forward Deployed Engineering ("the first FDE Manager in DACH"), a Technical Deployment Lead and an Applied AI Engineer, Enterprise (Anthropic job board).
- OpenAI lists a Forward Deployed Engineer – Munich (published 5 October 2026; hybrid with three office days per week, up to 50% travel, 5+ years of experience) as well as several applied AI roles in Munich (OpenAI job board).
- Databricks is looking for a Forward Deployed Engineer (Belgrade, Berlin, Munich), a Staff Forward Deployed Engineer (Berlin, Munich), a Manager, Forward Deployed Engineering and a Sr. Engagement Manager, FDE (both Munich). The posting states explicitly: "FDEs are billable" (Databricks, Forward Deployed Engineer).
Two observations follow. First: Anthropic is only just building the role in Germany – the manager position is framed as building from scratch ("building from zero to one"). Second: German is explicitly required for customer-facing roles. This suggests that the providers expect projects with German-speaking customers, not only with English-speaking corporate headquarters.
This snapshot says nothing about the total number of FDE positions in Germany. Start-ups and software companies also use the title, sometimes for roles closer to classic implementation or customer support.
Salary: what can be documented – and what cannot
The search query "forward deployed engineer salary" is common, but the data for Germany is thin. The postings by Anthropic, OpenAI and Databricks mentioned above contain no salary information; at OpenAI, the display of the salary for this position is explicitly switched off in the job data. Salary portals show averages based on a few very different listings – from start-ups to AI labs. They are not suitable as a benchmark, which is why we do not give a figure here.
For the US there are compensation reports with considerably higher figures; we put them into context in our article on the Palantir model, including the caveat that they mostly come from providers with their own market interest. They cannot be transferred to Germany. Anyone negotiating should orient themselves on concrete offers and on comparable senior roles in software engineering.
Skills profile: what the role requires
A fairly clear profile can be derived from the postings:
- Solid software engineering. Production-grade code in Python and at least one other language such as TypeScript or Java; frontend and backend.
- Production experience with LLMs. Not just API calls, but agent development, evaluation methods and operation at scale.
- Discovery with the customer. Finding out which problem is worth solving, understanding workflows, bringing domain experts and IT to the same table.
- Enterprise IT and security. Permissions, data protection, security review, integration into legacy systems.
- Dealing with ambiguity. Anthropic writes "thrive under ambiguity", OpenAI "make fast, sound decisions under pressure".
- Communication and language. Explaining technical matters to different audiences – in Germany often in German and English.
- Pattern recognition. Capturing repeatable solutions and feeding them back into one's own product.
Industry knowledge is a plus: Anthropic names financial services as well as healthcare and life sciences.
Ways into the role
There is no formal training to become a Forward Deployed Engineer in Germany. Typical routes that follow from the requirements:
- From software development: Developers who have already brought LLM applications into production and work close to business departments meet the technical core.
- From consulting and systems integration: Technical consultants with an implementation share bring customer experience; Anthropic explicitly names "Software Engineer with consulting experience".
- From pre-sales: Solutions engineers who want to build themselves instead of showing demos.
- As former founders: Anthropic explicitly encourages "former technical founders" to apply.
Learning offers such as the Claude Academy, which Anthropic describes as "Open to everyone", are open to individuals. Anthropic directs certifications and the Basecamp programme at partner organisations (Claude Frontier Academy, programme page). The most important proof, however, remains a running system that people actually use – with evaluation, monitoring and a clean handover.
Claude Frontier Academy: what Anthropic has announced
On 2 October 2026 Anthropic announced the Claude Frontier Academy. The following information comes from this announcement and the programme page (both accessed on 4 November 2026):
- Goal and scale: Anthropic is committing 100 million US dollars and aims to train 10,000 Frontier Deployed Engineers by the end of 2027. So Anthropic uses "Frontier" here, not "Forward" – in its own Munich job postings, however, the role is called "Forward Deployed Engineer".
- First cohorts: Engineers from Accenture, Bain, Capgemini, Commonwealth Bank of Australia, Deloitte, McKinsey, Morgan Stanley and Novo Nordisk, "and others".
- First programme: the Frontier Deployed Engineer Residency, modelled on medical training: learning from practitioners, practising on realistic cases, being assessed before working independently.
- Target group: hands-on software engineers with strong fundamentals, a record of building with LLMs, experience helping others adopt AI, and a habit of tackling their business's most important problems. Prior experience building AI agents is not required.
- Structure: First a multi-day in-person programme with Anthropic engineers and licensed instructors – according to the programme page, four days: three days building a Claude system for a simulated enterprise, from the first customer request through security review to handover, and on the fourth day a graded assessment on a new scenario. Those who pass earn the "Claude Resident Engineer" badge. This is followed by a twelve-week residency in which participants lead a real Claude use case at their own organisation, supported by Anthropic engineers. At the end there is another assessment; those who pass earn the "Claude Frontier Deployed Engineer" badge. Anthropic expects the first in early 2027.
- Access: by nomination only. Organisations nominate their strongest engineers, each with a named Claude project. Whether an organisation can take part is clarified with its Anthropic account team or its Partner Account Manager. According to the programme page, the residency is "in early access with a select group of customers and partners"; further programmes are to be announced in 2027.
- Locations: Cohorts are running in San Francisco, New York and London.
- Place in the partner ecosystem: The Academy builds on the Claude Partner Network, in which, according to Anthropic, professionals across 46,000 firms have earned more than 175,000 Claude certifications and nearly 4,000 people have completed the Basecamp programme.
What the pages do not say: Germany, Europe or the EU are not mentioned; the only European location is London. Participation fees are not stated. Applying as an individual is not foreseen.
What this means for Germany: For most companies here, the Academy is currently not a direct route. It is aimed at Anthropic customers and partners with their own account team, and the first cohorts consist mainly of large consultancies and corporations. It still has an indirect effect: when large consulting firms have engineers trained to a common standard, "FDE" becomes more comparable as a qualification – and the question of whether a provider actually builds or only advises becomes easier to ask.
What companies without their own FDE team can do
An in-house FDE team only pays off once several AI systems are being built and operated in parallel. For most mid-sized companies, that is not the starting point. The way of working can still be used:
- Choose a use case that matters. A process with measurable effort and clear ownership in the business department – not the most impressive demo case.
- Clarify access before anyone starts. Which systems, which data, which permissions, who signs off? Missing data access is one of the most common reasons pilots stall.
- Embed an external partner who builds. Check whether the partner works in your systems, writes code and is accountable for go-live. We have broken down the differences between implementation and pure consulting in our comparison of end-to-end vs. AI consulting.
- Plan the handover from the start. Documentation, operational responsibility and knowledge transfer to your team belong in the contract, not at the end.
- Name one person internally. Someone in the company should build along and then carry operations – an internal FDE on a small scale.
This is also how we work at Context Studios. We are a small AI consulting firm from Berlin: founder Michael Kerkhoff has been working AI-natively since 2024, as a UG since 2025, and brings in freelancers as needed. We work directly in our customers' systems, build ourselves and hand over so that your team can carry the result forward. Our Datura case study shows what that looks like in practice. We bill projects at a fixed price after scoping, with a proposal within 48 hours; to get started there are workshops at fixed prices of €1,500, €2,500 and €4,500. If you are looking in the capital region, you will find our local approach under AI consulting Berlin. We are not part of the Frontier Academy.
FAQ
What does a Forward Deployed Engineer do? An FDE works embedded with the customer and takes an AI system from problem definition through prototype and security review into production. The FDE writes code in the customer's environment and is measured by actual use.
Are there Forward Deployed Engineer jobs in Germany? Yes. As of 4 November 2026, Anthropic (Munich), OpenAI (Munich) and Databricks (Berlin, Munich), among others, are advertising FDE positions. Anthropic requires native German for its Munich FDE position.
How much does a Forward Deployed Engineer earn in Germany? We did not find a reliable salary survey for Germany. The postings by the large providers state no salary, and averages from salary portals are based on a few very different listings.
Can I apply to the Claude Frontier Academy? Not as an individual. Participation is by nomination from one's own organisation; whether it is eligible is clarified with its Anthropic account team or Partner Account Manager. Cohorts are so far running in San Francisco, New York and London.
Do mid-sized companies need their own FDE team? Usually not at the start. A narrowly scoped use case with an external partner who builds and hands over cleanly makes more sense, plus one internal person who carries operations afterwards.
Sources
- Anthropic, Claude Frontier Academy: $100M to train 10,000 engineers, 2 October 2026, accessed 4 November 2026 – Link
- Claude, Claude Frontier Academy (programme page), accessed 4 November 2026 – Link
- Anthropic, job posting Forward Deployed Engineer, Munich, accessed 4 November 2026 – Link
- Anthropic, job posting Manager, Forward Deployed Engineering, Munich, accessed 4 November 2026 – Link
- Anthropic, job posting Applied AI Engineer, Enterprise, Munich, accessed 4 November 2026 – Link
- OpenAI, job posting Forward Deployed Engineer – Munich, published 5 October 2026, accessed 4 November 2026 – Link
- Databricks, job posting Forward Deployed Engineer (Belgrade, Berlin, Munich), accessed 4 November 2026 – Link
- Claude Academy, accessed 4 November 2026 – Link