According to Business Insider statistics released in April 2026, job postings for forward-deployed engineers (FDEs) have risen roughly 729% year over year. The increase points to the overwhelming demand for talent in a role that combines software engineering with the hands-on work of making technology function well inside a customer’s business. However, the increased demand also offers a window into a challenge faced by enterprise AI: building a working demo is not the same as building a system that businesses can rely on long term. [Text Wrapping Break][Text Wrapping Break]Sanjay Mehta, Global Operations Leader of Blackstraw, a global AI consulting and digital transformation firm, describes their hiring priorities as shifting attention toward deployment. “We are working on helping businesses integrate AI into existing operations, maintenance after launch, and assigning responsibility when something goes wrong,” Mehta explains. This one employer’s experience speaks to the market by helping explain the work behind the growing demand for forward-deployed engineers.
Closing gaps between prototype and a fully functional system
“Earlier enterprise AI projects often focused on exploration,” says Mehta. Teams were tasked with testing use cases, developing proofs of concept, and demonstrating what would justify further investment. However, today’s projects may involve a different set of demands. [Text Wrapping Break][Text Wrapping Break]“Technical problems differ between enterprises,” says Mehta. In his experience with Blackstraw, his teams have seen enterprises want AI agents to take actions within SAP-driven operations without disrupting existing business rules and governance. They helped a business that needed to address slow releases and limited visibility with machine learning models running across multiple environments. A third client had moved AI agents into full deployment, only to find their operating costs increasingly difficult to track and predict.[Text Wrapping Break][Text Wrapping Break]Mehta and other AI consulting firm leaders have seen a central theme to these various issues: answering the question of “Who owns the project once it is live?” [Text Wrapping Break][Text Wrapping Break]The question changes the skills that employers look for and need from their potential employees. While a prototype can demonstrate capability, a production system has to work with the organization’s data, fit its existing processes, and remain manageable after initial development. The people answering the call for the open roles for FDEs must consider this shift in the market. They sit at the intersection of writing code and making it function in the real world for customers. [Text Wrapping Break][Text Wrapping Break]The distinction is important as the job postings for FDEs continue to grow. The job is not simply to explain technology to a customer or pass requirements back to a development team. Today’s FDEs must combine an understanding of the business problem with the ability to take responsibility for implementing the technical solution.
The less visible work is becoming more important
While forward-deployed engineering hiring is having a moment of growth, leaders in the space like Mehta highlight the need for infrastructure to keep AI systems running. [Text Wrapping Break][Text Wrapping Break]“Machine-learning operations cover the behind-the-scenes work of deploying, tracking, and maintaining machine learning models,” Mehta explains. “Introducing a machine-learning operations framework can reduce model deployment time by a significant amount. Nobody posts a demo of a well-versioned model pipeline, yet that pipeline decides whether a model runs for a week or for years.”[Text Wrapping Break][Text Wrapping Break]Even the familiar roles have changed shape. From GenAI engineers who can ship RAG systems and agents to real users — and tell you what is broken — to business analysts who can take a messy business problem into something a model can actually do, the job titles have stayed the same, but the job descriptions have changed significantly.
Hiring for the new era of AI work
As leaders like Mehta have discovered, keyword filters used during the hiring process can only take you so far. “AI skills now appear on a large share of resumes, so the keywords alone don’t tell us much about how someone approaches real work,” Mehta says. [Text Wrapping Break][Text Wrapping Break]What hiring teams are discovering is that handing a potential employee a problem, letting them build, and asking them to explain their choices can be a more effective way of analyzing one’s skills. “The answers show us how someone thinks through a problem, which matters more to us than any single tool on their resume,” says Mehta.[Text Wrapping Break][Text Wrapping Break]Hiring departments are also turning to their current stable of talent, relying on the belief that when someone on their team recommends a former colleague, they’re putting their reputation on it, and they know exactly what the job involves.[Text Wrapping Break][Text Wrapping Break]Businesses also have an eye on retention. Leaders are coming to realize that top talent is more likely to stay with a company if they have access to real client work early on. They also need to be given space to learn within a field that can radically change every few months. As teams work together and grow, top employees who prove to be eager to learn and remain flexible will have more opportunities to step into bigger roles. [Text Wrapping Break][Text Wrapping Break]By looking closely at the new listings for open roles in the AI space, you can learn a lot about a company’s view of AI. Today’s forward-thinking businesses are looking for professionals who can deploy AI, run it, and own it once it is live. Where many of these job descriptions wouldn’t have made sense even a few years ago, the landscape has shifted dramatically. The market is now telling employers and job-seekers what is needed, and both parties are ready to listen.