AI adoption is changing the way technology teams are organised. Traditionally, responsibilities were clearly divided. Business teams defined their needs. Product teams designed the solution. Engineers built it. AI is changing this process.
The main challenge is now integrating generative AI models into existing IT systems. These models can produce different results from the same input, while traditional systems rely on strict rules and predictable outcomes.
Writing code alone is no longer enough to solve this challenge. New roles are becoming more common, including Forward Deployed Engineer (FDE), Product Engineer, Product Builder and AI Engineer.
These roles reflect a wider shift in software development. Businesses increasingly need professionals who combine technical skills, product knowledge and the ability to understand business challenges.
From AI Engineer to Forward Deployed Engineer
To build an effective AI delivery team, businesses need to understand the difference between developing AI models and integrating them into real systems.
AI Engineer: The model specialist
An AI Engineer, sometimes called a Machine Learning Engineer, focuses on the technology behind AI models.
Their work includes training models, adjusting model weights, fine-tuning performance and reducing inference latency.
They usually work in research and development or controlled technical environments.
Their main goal is to improve how the AI model itself performs.
Forward Deployed Engineer: Connecting technology with business needs
Demand for Forward Deployed Engineers has reportedly increased by 800% in just a few months, adding pressure to an already competitive talent market. Yet the FDE role is not new. Palantir introduced this model in the 2010s to integrate complex data analysis software into highly secure government and financial systems.
Why is it gaining attention again? AI solutions are becoming more complex. Traditional IT service providers sometimes struggle to deploy tailored solutions quickly enough. As a result, AI model providers are developing their own integration teams, following an approach similar to Palantir's model and forcing the entire market to adapt.
Recent examples include OpenAI's Deployment Company, Microsoft's Frontier Company initiative and the joint work between Accenture and Google Cloud to deploy Gemini Enterprise.
The FDE role covers three main stages:
- Discovery and technical assessment: Review the client's data, processes and existing systems.
- System design: Design the solution architecture, such as a Retrieval-Augmented Generation (RAG) pipeline, within the existing IT environment.
- Development and deployment: Build the integration, secure data flows and deploy the solution into production.
An FDE combines software engineering with a practical understanding of the client's business.
Product Roles: Product Engineer, Product Builder and FDPM
While FDEs focus on deploying external technology into client systems, new hybrid roles are also emerging in internal product development. A Product Engineer is a software developer who does more than implement technical specifications.
They take part in product decisions, consider the user experience and assess how technical choices affect business results. Their work connects software architecture with product performance.
The rise of No-Code and Low-Code tools has created another role: the Product Builder. Product Builders focus on speed. They combine existing tools and technologies to test a business idea in just a few days. AI is now changing both roles.
Product Managers can use AI tools to create technical prototypes without relying entirely on development teams. The main difference lies in the goal.
A Product Builder creates something to test an idea quickly, often through a proof of concept (POC) or minimum viable product (MVP). A Product Engineer builds software designed to remain reliable as it grows, with the architecture needed for production.
In practice, these technical specialists work alongside a Forward-Deployed Product Manager (FDPM). The FDPM manages the strategic relationship with the client. They ensure that engineering efforts address real business needs rather than focusing solely on technical complexity.
Three roles, three different approaches
The FDE, Product Engineer and Product Builder come from different professional backgrounds. The FDE has roots in software vendors and consulting. The Product Engineer comes from software engineering. The Product Builder comes from product development.
However, their responsibilities increasingly overlap. All three can take part in discovery, development, deployment and performance measurement. AI is making the boundaries between these roles less rigid. In 2026, an FDE may combine platform expertise with direct knowledge of a client's operations.
A Product Builder may take on tasks traditionally handled by Product Managers, designers and developers. Yet these roles remain distinct because they serve different stages of a project. A Product Builder helps test an idea quickly. A Product Engineer builds and maintains reliable software in production. An FDE integrates a technology provider's solution into a complex IT environment, often involving several disconnected systems.
Comparing hybrid AI toles
AI Delivery at Novity: Combining skills in practice
Understanding these roles is the first step towards successful technology integration. At Novity, combining different skill sets is not just a theoretical concept. It reflects how we work in practice.
We support clients with teams that bring together software engineers, product specialists and professionals who can assess and challenge business needs.
In recent AI projects, our engineers have naturally taken on responsibilities similar to those of Forward Deployed Engineers. They work directly within the client's environment to assess requirements, design the architecture and support secure deployment into production.
The skills we bring depend on the client's organisation and the stage of the project.
For example:
- To test a business idea, we focus on product skills and rapid development.
- To integrate an AI agent into an existing ERP system, we prioritise the engineering and integration skills of an FDE.
Our teams also benefit from continuous internal training and recognised certifications, including Claude Certified Architect – Foundations (CCA-F). Novity holds Certified status in the Claude Partner Network. This strengthens our ability to apply Anthropic's security and architecture standards to client projects.
About Anthropic's Claude Partner Network
The Claude Partner Network (CPN) is how Anthropic works with the firms that put Claude into production for customers.
The Services Track recognizes partners for building a certified delivery team, delivering Claude use cases for clients, and proving the work through public customer references.
Partners advance through Certified, Select, Preferred, and Global Premier tiers.
Sources
- ZDNET - AI in business: AI engineer or Forward Deployed Engineer – which delivers greater value?
- Les Échos - AI: “Finding the talent capable of turning promising technology into game-changing solutions: the role of the Forward Deployed Engineer”
- CIO - OpenAI’s new AI consulting offering raises questions about trust and strategy
