Companies typically face challenges in recruiting candidates who are skilled in AI, machine learning and generative artificial intelligence though they have hundreds of vacancies.
Neuron7.ai has a problem that is 'painful for sure' when it is hiring artificial intelligence engineers.
The resolution intelligence company says for 100 applications, it is able to hire three people at the most.
The company, which is based in the United States and has offices and employees in India, provides resolutions that include helping its clients improve prediction for how software parts will behave, errors that might come up, onboarding customers and gathering information.
"As a Java developer if you cannot answer what is object-related transforming, then it is a big red flag for us because it gives me the impression that you have used Java application but have not updated it or you may have used it intermittently but have not gone deeper," said Amit Verma, head of AI and engineering at Neuron7.ai, referring to how data is converted from one object type to another on the widely used software platform.
Companies typically face challenges in recruiting candidates who are skilled in AI, machine learning and generative artificial intelligence though they have hundreds of vacancies.
The problem is not just that there are more openings than there are qualified people, but also that there is a difference between what companies need as essential skills and what candidates bring to the table.
Open positions
"The current active demand for AI-focused specialist talent is a little under 2,000 openings and there just isn't enough accessible talent in the market for these roles," said Kamal Karanth, co-founder of specialist staffing company Xpheno.
"The overall active demand for AI-linked roles, including the services spectrum, is more than 20,000.
"The challenge is that the right talent for these roles are either getting built or not available and inaccessible.
"These conditions have driven a significant demand-supply gap in the talent market for experienced AI skills," he said.
The problem means work at enterprises takes longer at a time when the world of AI and GenAI is changing rapidly and the realms of what is possible in that ecosystem is being redrawn every quarter practically.
According to experts, the focus on AI is only a decade old and talent for the technology is being built gradually.
"What we are looking for does not exist in terms of skill sets," said Raj Gopalakrishnan, CEO, Kogo AI.
"What we do is we hire the closest match by looking at candidates whose logical thinking quotient is very high and who understand data logic and engineering from a broader building point of view."
The hiring challenge comes at a time when companies are paying eye-popping salaries for AI and data science jobs compared to engineers with traditional software skills.
AI companies are ready to open their wallets for the right talent.
A recent report by staffing company Quess Corp said that specialised roles in cybersecurity, GenAI, platform engineering, and user interface and user experience can command as much as Rs 50 lakh per annum with just eight years of experience in Bengaluru.
Some key roles that have gained prominence over the last decade are engineers skilled in deploying large AI algorithms and natural language processing that is used to build chatbots.
There is also demand for engineers who can provide AI and ML solutions.
"Prompt engineering will only take you so far and people who have used it and done this work will also tag themselves as GenAI or large language mode experts," said Verma.
"But when you talk to them, ask about transformer architecture, what is the difference between GPT2 and GPT3, why is the architecture evolving, what are the limitations, how and when to train those models, that is where the real problem creeps in," said Verma, talking about various AI programmes and tools.
Aditya Narayan Mishra, managing director and CEO of Ciel HR, said candidates may list skills on their resumes but their experience is often limited.
"This is especially evident in fast-evolving domains like AI and cybersecurity, where staying on top of emerging technologies such as LLMs, zero trust, and SOAR (security, orchestration, automation and response) is critical."
Skilling matters
"Signing up people in these [AI] domains is difficult and it mainly depends on upskilling and reskilling.
"If we are looking for someone with five years of experience in LLMs, it is not possible.
"Rather, we look for people with problem-solving and analytical skills, and who are ready to learn and a cultural fit," said Savita Hortikar, vice-president-global head of talent acquisitions, at Fractal Analytics.
Fractal's analysis of applications show that if a company is looking to fill up a position that requires seven to eight years of experience, almost half of the candidates are 'freshers or ones with two to three years of experience.
"Most of them would have done some online certification courses or some specialisation in colleges. Finding the right fit in this field is always hard now. We have huge volumes coming in and we need to go through the details," Hortikar said.
"In India talent is not the problem but it is the quality and experience which is lacking," she added.
Xpheno's research shows that applying experience and expertise filters brings the active pool of senior AI engineering talent in India to just a little over 3,000.
The wider pool of engineers experienced in AI tools is under 18,000 in India.
There are more engineers who specialise in AI services than who have core development capabilities for the technology.
"The current low volume of experienced AI talent in India and the velocity with which higher-end AI upskilling is happening will keep the demand-supply gap open for two to three years to come," said Karanth of Xpheno.
"Demand for AI-related specialist talent is expected to intensify as enterprise AI use-cases increase and integrate into everyday enterprise processes and tech consumption.
"Estimates are that by 2030, AI jobs have a potential to multiple eight-fold to 10-fold the volume that we are currently witnessing.
"India's AI talent build and supply should increase and speed up to meet this demand in time to ride the wave and not let it pass," he said.
Feature Presentation: Ashish Narsale/Rediff
© 2025