As artificial intelligence changes how companies develop software and deploy technology, demand for traditional application-development roles is shifting toward AI engineering and multidisciplinary jobs that connect AI systems with enterprise operations. For technology workers, the challenge is to build the skills employers increasingly expect; for India, it is an opportunity to capture new work as global capability centres expand.
Vijay Swaminathan, CEO of talent intelligence firm Draup, discusses why technology hiring is changing, how AI could create new employment opportunities, and why forward-deployed engineers could become an important growth area. He also explains what the shift means for entry-level engineers and the skills they will need to build their careers.
The interview has been edited for clarity and length.
Cybernetix: Technology workers are worried that AI will take away jobs. How do you see this playing out, particularly in India?
Vijay Swaminathan: A lot of the job losses in India are happening on the IT services side, particularly in traditional application development. But there is a compensating force: global capability centres (GCCs) are expanding and hiring. It is not a one-to-one replacement, where every job lost in IT services is matched by a job at a GCC.
The qualification process has also become more stringent. Employers want strong engineering skills, but they increasingly expect people to demonstrate deeper expertise. Someone who loses a job as a traditional application developer cannot simply apply for an AI engineering position and expect to get it without showing the necessary skills.
There is a period of mismatch between the skills people have and what employers need. That can be mistaken for a lack of demand, when the underlying issue is often that candidates are not yet qualified for the roles being offered.
Cybernetix: How much of the recent reduction in technology employment is actually due to AI?
Swaminathan: Two things are happening. During COVID, technology companies expanded their workforces rapidly and subsequently corrected their hiring. Some of those reductions would have happened even without AI, but people tend to associate events that occur around the same time.
At the same time, companies are investing heavily in AI and the infrastructure needed to run it. They are reviewing their recruitment and support functions to see where AI can help them do more with fewer people. That adjustment is happening, but it does not mean demand for core technology talent has disappeared.
India has around 2,000 GCCs, and there is substantial room for further growth if the country can supply the skills these centres need. The priority should be rapid reskilling and upskilling, involving both employers and individuals.
Cybernetix: Which technology roles do you expect to see growing as companies adopt AI?
Swaminathan: There are two broad areas. One is AI builders, including AI engineers and architects. The other is forward-deployed engineers, or FDEs, who bring together multiple skills.
An FDE needs some knowledge of AI, data engineering and software engineering, along with an understanding of the business domain. These people help translate technology into working systems for an enterprise.
We have many AI models, and companies are experimenting with them. But building a prototype is different from deploying it at scale inside an enterprise. A developer might use an AI tool to build an application relatively easily, yet still not know how to integrate it into enterprise systems and implement it at scale.
That is why I believe forward-deployed engineering will be a major area of growth. I expect demand for this talent eventually to exceed demand for core AI builders, although I would not put a percentage on that.
India is well positioned because it has a large base of software engineering talent that can build on its existing skills in AI and data engineering. FDEs also need domain knowledge and the ability to work across disciplines, rather than only highly specialised AI research skills.
Cybernetix: Draup’s job-posting analysis identifies several categories of AI-related talent. What does the distribution tell us?
Swaminathan: In our analysis of job postings, AI builders account for 27% of the roles, infrastructure for 20.5%, forward-deployed engineers for 18.5%, guardrail-related roles for 15%, user experience and design for 11.5%, and technical support for 7.5%.
These categories reflect different requirements around building, running, deploying and supporting AI systems. The demand is not confined to people who develop AI models. Companies also need the infrastructure to operate them, people who can deploy them in business environments, and capabilities to manage the associated safeguards.
Note: The percentages represent the categories in Draup’s job-posting analysis, not the distribution of all technology jobs.
Cybernetix: You mentioned a 15% to 20% reduction in entry-level jobs at Fortune 500-type companies. What is driving that, and what does it mean for new graduates?
Swaminathan: I have heard this anecdotally; I do not have a specific dataset for that estimate. One reason is that some tasks traditionally given to entry-level employees, such as basic checks and quality-related work, can now be automated using AI. Managers are also reconsidering what work they can assign to junior employees.
The entry point has become more difficult, but it has not closed completely. Internships, contract work and project-based opportunities can help graduates gain practical experience. They may not offer the same compensation or the immediate full-time employment that many graduates expected a few years ago, so there needs to be an adjustment in expectations.
I would encourage new engineers to treat the years after graduation as a period of continued learning. That could mean additional courses, internships, contract assignments and projects that build hard skills, particularly in AI-related areas. The early stages of career growth may be slower, but engineers who continue developing their skills can improve their prospects over time.
Cybernetix: What should technology professionals do if their existing skills no longer match the roles employers are hiring for?
Swaminathan: They need to identify the skills that are in demand and make a deliberate effort to acquire them. An application developer cannot assume that experience in one technology automatically qualifies them for an AI engineering role.
Practical experience matters. Completing projects, taking internships or working on contracts can help someone demonstrate that they can apply their skills. Employers also have a role in reskilling their workforces, but individuals need to take responsibility for their own learning.
The important point is to build depth rather than rely only on a credential or a job title. If someone can demonstrate the skills a company needs, the opportunities can be very different from those available to someone who has not made that transition.
Cybernetix: Beyond the large technology companies announcing layoffs, where do you see opportunities for new jobs?
Swaminathan: We tend to focus on a handful of well-known companies when they announce layoffs, but we also need to look at companies and industries that are emerging around AI.
AI could enable new businesses and applications in areas such as medical devices and materials discovery. These opportunities may create jobs in companies that are not yet as prominent as the established technology giants. Not everyone needs to become an entrepreneur; people can also join smaller companies and help build them.
I am optimistic about the potential for AI to create new employment, although I cannot give a reliable global figure for how many jobs will be created or how those jobs will be distributed across countries and industries. The nature of technology employment is changing, and people may need to look beyond the traditional expectation of joining a large IT services company directly from campus.
Cybernetix: What should CIOs do to prepare their organisations for this shift?
Swaminathan: CIOs need to think about the data foundation, organisational structure and learning culture. Enterprises may be able to build AI prototypes, but deploying them at scale requires the right data and the ability to integrate the systems into existing operations.
Organisations also need to reconsider how teams work together. A more horizontal structure can help people with different skills collaborate on business problems. And learning cannot be treated as a one-time activity; employees need opportunities to keep building their capabilities as technology changes.
CIOs should not look at AI only through the lens of reducing headcount. If AI enables a team to do more work with fewer people, the organisation should also consider what additional work it can take on and what new services or capabilities it can deliver.
