There is an emerging role that is in demand like never before. One for which there is no standard education, no fixed background required, and no clear-cut CV needed.
The Forward Deployed Engineer.
FDE vacancies increased by more than 800 percent in 2025. OpenAI's own FDE team grew from 2 to 52 engineers in one year. Deloitte launched an entire service around it. And in February 2026, OpenAI announced multi-year partnerships with McKinsey, BCG, Accenture, and Capgemini, where OpenAI's FDE teams work directly with consultancy teams in client projects.
The reason is simple. Raw model access is no longer sufficient to deliver value. What organizations need are people who connect technology and business; who understand what problem is worth solving and who build a solution that actually works in the client's practice.
As an interim professional, you probably already have exactly that combination. You just might not know it yet.
What an FDE actually does
A Forward Deployed Engineer works directly in the client's practice. Not from the headquarters, but close to the operation. The term is no coincidence: forward deployed refers to people deployed close to the action.
A concrete example. A logistics company processes thousands of emails daily. An FDE first shadows the planners. Not to sketch a solution, but to understand how the process really works, where the variation lies, and what quality means in this specific context. Only then does the FDE determine if and how AI adds value. A solution is built that automatically categorizes emails, integrated into the existing workflow; then it is measured whether the result actually delivers value.
The latter is the most underestimated part. Many AI projects do not fail on technology, but because the wrong problem is being solved. The FDE is the one who prevents that.
Where consultant, business analyst, product owner, and software engineer are traditionally four separate roles, an FDE brings these disciplines together in one profile. With every handover between roles, context is lost. An FDE prevents those translation losses; the FDE is the continuous bearer of the judgment about the problem, the solution, and the moment when the two actually align.
There is no standard profile
The common assumption is that you must be a classic software engineer to become an FDE. Practice shows otherwise.
Many successful FDEs have a hybrid background. Some started as consultants, others came from data, product management, or operations. What distinguishes them is not their education, but two things: the ability to quickly acquire new knowledge and the ability to connect disciplines in service of a concrete outcome.
A full-stack developer learns to build AI solutions. A product engineer delves into LLMs and agents. A data professional takes on more business responsibility. A consultant develops technical skills. Different routes, the same destination.
What you probably already have
Many professionals underestimate what they bring.
A consultant understands stakeholders and knows how organizations make decisions. A product owner understands user problems and the trade-offs between options. A full-stack developer understands how systems are put together and where assumptions break. A data engineer understands data flows and knows when a model is correct and when it is misleading.
These are not trivialities. That is exactly the profile organizations are looking for. The challenge is not to start over. The challenge lies in combining.
Which skills are crucial
Technical. AI tooling, LLMs, agents, automation, API integrations, cloud platforms. You don't have to be an expert in everything, but you need to understand how these components come together in a working solution. And you must recognize when an AI output operates outside its area of competence. The latter is not a technical skill. It is a judgment skill.
Business. Stakeholder management, process analysis, value creation, product thinking, problem definition. The question "what is success here?" sounds simple. In practice, it is the question that most AI projects do not explicitly ask and where they get stuck.
Personal. Ownership, curiosity, learning ability, communication. The technology changes too quickly to stand still. Learning ability is not a soft skill in this market. It is a competitive advantage.
From product manager to Forward Deployed Engineer
Lotte worked as a product manager at a medium-sized software supplier. No technical background, no mapped-out path towards AI.
She decided to change her course. Not through a formal program, but by experimenting outside working hours. Building small AI tools for problems she encountered daily. Hobby projects that never became large but taught her how models work, how to connect APIs, how to deploy something. And importantly: how to recognize when a model does something right and when it sounds convincing but is wrong.
Two years later, she combines product thinking with technical execution. She understands what users need, can build it herself, and knows when the AI output does not replace her judgment but actually asks for it.
What distinguishes her is not her CV. It is what she demonstrates.
What you can do today
Don't wait for the perfect assignment. Build an AI agent. Automate a process. Integrate an LLM with an existing application. Solve a real problem; then ask yourself the question most people skip: is this the right solution for the right problem? And how do I know that?
Look for assignments where you operate closer to the business than you are used to. Build a portfolio that proves delivery but also understanding of the context in which you deliver. Because this role is not about certificates or job titles.
It is about the ability to understand a problem, build a solution, and make that solution actually work.
The question is shifting
The biggest opportunities in the AI market do not arise for those with the most access to models. They arise for professionals who connect technology and business; who know what to do with those models and what not to do.
The market has now confirmed that. OpenAI, Deloitte, McKinsey, BCG, Accenture: they are all investing in the same profile.
That is why the Forward Deployed Engineer is so relevant. Not as hype, but as direction.
This article is based on a combination of public research, market analyses, and publications from OpenAI, Deloitte, BCG, McKinsey, Accenture, MIT/NANDA, Forbes, Fortune, CNBC, and Indeed/Financial Times, supplemented with scientific research into AI adoption and analyses by NXTminds on the rise of the Forward Deployed Engineer.
