A former software engineer turned chief people officer makes the case for a new kind of technical talent in HR
HR teams have traditionally relied on IT to build the technology they need or bought software off the shelf. Now, as AI makes it easier to create custom tools, some people teams are starting to build their own systems, too.
The shift could change the technical skills HR needs and how companies resource their people functions. Stanford’s 2026 AI Index found that 88% of organizations worldwide used AI in 2025, up from 55% in 2023. For some HR teams, the next step is moving beyond using AI tools to hiring people who can build them.
Gabriella Garcia, chief people officer and chief operating officer at Brooklyn-based payroll and HR software company Warp, is one example of that shift. She came to her role through engineering rather than HR and now oversees six teams, from growth and marketing partnerships to people and talent.
“About two years ago, it became pretty obvious, with the rate that it was becoming easier to produce software, that maybe we should be hiring engineers or even allocating internal resources, whether that’s an actual engineer on the people team or getting capacity from a few engineers,” Garcia said.
From software engineer to people leader
Originally from Bogotá, Colombia, Garcia grew up in Miami and double majored in mechanical engineering and computer science at MIT. Her career took her through engineering roles at Facebook and Apple, product management at Google, and finance at Two Sigma, a quantitative hedge fund, and an early-stage venture capital fund.
She joined Warp as one of its first operational hires, when the company had about 10 people. Today, it has 76. Her initial focus was building a recruiting funnel and engine.
“That initially started this waterfall of owning a bunch of HR and people initiatives, and really just gaining so much love and respect for the industry and the role in general,” Garcia said.
She said the move also came with respect for what she didn’t know.
“I’m very fortunate to have assembled a pretty great group of mentors and advisors and overall a community in this space,” Garcia said.
She still draws on her engineering training, though, including a preference for simple systems.
“Another aspect that’s helped me a lot, and this is surprisingly reflective of how a lot of chief people officers scale, is this concept of elegance in which the simplest system is usually the best and the most scalable,” Garcia said.
The rise of the people engineer
Garcia said people teams began adding data scientists about eight to 10 years ago, and engineers more recently.
“I think the shift that I’m very excited about, and I’ve written about it as well, is that next phase of that, which is the rise of the people engineer,” Garcia said.
Garcia said a company can fill that role in two ways. It can hire an engineer onto the people team or upskill someone already there, such as a people operations employee or people data scientist, into a full engineer.
Garcia said Shopify, whose CEO is an angel investor in Warp, has a 50-person engineering team for its people function, while Lovable’s chief people officer has been upskilling people operations staff and retitling them as people engineers.
“They have essentially crossed that boundary of just playing with Claude to really building scalable systems,” she said.
Technical skills are climbing beyond HR, too. The World Economic Forum’s Future of Jobs Report 2025 projected AI and big data, networks and cybersecurity, and technological literacy as the three fastest-growing skills through 2030. Gartner also expects 30% of organizations to form blended HR and IT teams by 2029, according to HRD America reporting.
Building HR tools from the inside
Warp’s people team builds tools for its own work. During open enrollment, for example, two members of Garcia’s team have been building an interactive benefits guide drawing on each plan’s summary of benefits and coverage and the reasoning behind the company’s benefits choices.
Garcia, who has 14 direct reports, also built a tool for Warp’s biannual performance reviews.
“I built a Claude bot that I gave to each of my managers that connects to Notion, Google Docs and their meeting recorder systems,” she said. “It would then parse through the last six months of one-on-ones and start auto-filling some of the questions or even data points.”
Using the company’s scoring rubric, the bot also suggested examples to input in each section of the review, she said.
Garcia said the bot helps with recall. With so many direct reports, she said, it’s hard to remember what each person has accomplished, and it can pull that context from one-on-one notes.
She also thinks about how to keep people from simply accepting what it drafts.
“One thing that I think about a lot is how do I make people not lazy and not just accept those draft performance review questions,” Garcia said. “A lot of the purpose of a performance review is to be introspective and to be really thoughtful.”
Two researchers who study workplace behavior made a similar point to HRD America, describing AI as a starting point for feedback rather than a replacement for managers.
Garcia said AI strategy should be a cross-functional effort driven by the people team, with IT owning implementation because of the personally identifiable and protected health information that goes into AI systems. A Clutch report covered by HRD America found that almost half of employee-built agents can access HR information. Finance, she added, should own return on investment.
Asked whether other companies would benefit from hiring engineers to build AI tools for HR, Garcia said yes, and described several ways to approach it. Companies can upskill non-technical employees to use AI, adopt AI-native tools, or hire engineers, which she noted are expensive. The right fit depends on budget, prioritization and risk tolerance, she said, and smaller budgets shouldn’t rule out building.
“What you get is a spectrum of what is the right fit for my company today,” Garcia said. “So I would say start small. I call it don’t crawl, walk, run. Do a crawl and then a run.”
