When Kathleen Pearson was seeking ways to improve how managers conduct performance evaluations in her organization, her first thought wasn’t to put those leaders through a traditional training program. Instead, Pearson, CHRO of global law firm McDermott Will & Emery in Los Angeles, opted to create a custom generative pre-trained transformer (GPT) to help enhance those managerial skills — a tailored artificial intelligence assistant built using the firm’s enterprise ChatGPT account.
Pearson believed creating such a digital performance support tool — which managers could access in the moment of their learning need — would be a more effective way to elevate managers’ skills than acquiring knowledge through training courses that can quickly dissipate with the passage of time.
“The custom GPT helps managers write performance evaluations based on HR guidance and best practices,” Pearson said. “It mirrors our performance evaluation process and gives managers a more structured way to write their reviews. Like many managers, those in our law firm often find it difficult to write meaningful and objective evaluations.”
The custom GPT, which stores prompts and documents that can be used repeatedly, has another important benefit: It significantly reduces the time Pearson’s HR team has to spend auditing performance reviews throughout the year to ensure they meet organizational standards.
“The custom GPT reduces what would have been weeks’ worth of auditing work to a few days,” Pearson said. “By reducing the administrative burden on my HR team, the tool also frees them up to make better connections and build rapport with business unit leaders, bringing more of the human touch back to HR.”
Rise of Citizen Developers in HR
Pearson is one of a growing number of HR leaders spearheading a movement toward “do-it-yourself” chatbot and software application creation in the HR function. Dubbed “citizen developers” by technology industry analysts, these HR professionals and other employees outside of IT departments are increasingly using no-code and low-code software development platforms along with generative AI systems to build tools that automate a variety of manual HR processes. Their efforts often save valuable time and make the HR function more productive and efficient.
Building custom AI assistants and software apps offers multiple benefits for HR departments. It reduces the need for HR to wait on IT to build such tools and can save the expense of purchasing off-the-shelf apps from third-party vendors.
“The advantages include faster speed to execution and meaningful productivity improvements,” said Hiten Sheth, a director analyst with Gartner’s HR practice. “Citizen developers can reduce HR’s dependency on IT and have the additional benefit of fostering a culture of innovation within human resources.”
HR professionals are building custom AI assistants that include not only GPTs from OpenAI’s ChatGPT but also similar custom AI tools created on Google’s Gemini and Anthropic’s Claude platforms. They’re also creating software apps for a variety of tasks such as creating or modifying careers sites, automating the delivery of offer letters to job candidates, teaching employees prompt engineering skills, building custom reports that track employee turnover, and helping managers role-play difficult employee conversations.
A 2023 Gartner study projected that the number of users of low-code applications who aren’t in IT departments will grow from 60% in 2021 to 80% in 2026. Research from Forrester also found the AI-driven growth in citizen developers and low-code development tools will push the market for these do-it-yourself platforms to approximately $30 billion by 2028, a 21% growth rate from a $13 billion market at the end of 2023.
Some experts believe it will soon be a common expectation for employees in many departments, not just those in IT, to use low-code and no-code development tools to create apps or custom AI assistants in the same way they now use email, spreadsheets, or presentation deck software. Low-code and no-code platforms feature intuitive interfaces, drag-and-drop tools, and pre-defined templates that make it easier for nontechnical employees to build software applications.
Custom GPTs for Interview Questions, Difficult Conversations, and Benefits Information
Another HR executive who’s a proponent of creating custom AI assistants is Chris Courneen, senior vice president and global head of human resources for MS International Inc. (MSI), a supplier of flooring and countertops based in Orange, California.
Courneen has built three custom GPTs and is now teaching his HR staff how to build the tools as part of a formal learning program on the LinkedIn Learning platform.
The first GPT Courneen created was designed to help hiring managers ask more tailored, behavior-based questions in job interviews. Managers at MSI were initially struggling to ask good questions, but this GPT gives them a structured, user-friendly way to do it. Courneen said they now simply upload a job description and the candidate’s resume into the custom GPT, the system analyzes both documents, and AI suggests specific traits and skills to assess.
The custom GPT then generates a list of behavior-based questions, each with an explanation of why it was chosen and follow-up questions to ask, Courneen explained. The tool also allows managers to role-play interviews, with GPT taking the role of the candidate.
Courneen’s second custom GPT was designed to help MSI managers navigate difficult conversations with employees.
“It mimics scenarios where workers aren’t receiving feedback well, allowing managers to practice in those situations,” Courneen said. “Unlike what ChatGPT will do by default, we’ve built our custom version to create pushback on the manager during the role-play —and challenge them to think through how to respond when employees say certain things or become overly defensive.”
Courneen’s third, soon-to-be-released custom GPT will answer questions employees have about their medical, dental, vision, and supplemental insurance packages.
“It’s our new health insurance expert,” Courneen said. “It’s been trained on virtually everything related to our company insurance and also all the technical documents employees are often hesitant to read.”
This benefits-related GPT will sit on MSI’s benefits website behind a firewall because of the sensitive nature of the data. Courneen said that while it took a fair amount of time to create the tool, it didn’t require advanced technical knowledge.
“Most of the work in building the benefits GPT came in organizing the data that feeds into it,” he explained. “The low- code element of it made coding the GPT the easier part of the project.”
Teaching the HR Team to Build AI Assistants
As HR leaders find more use cases for custom AI assistants, many are encouraging their HR teams to look for opportunities to build them, as well. Some have developed formal training programs to teach their teams how to craft these tools.
On LinkedIn Learning, for example, Courneen created an “AI in HR” learning path that teaches HR business partners on his team how to create custom GPTs as one part of the curriculum.
Pearson is also teaching her HR staff how to build custom GPTs.
“I have 10 direct reports, and all of those directors are learning how to build their own GPTs and share them within their groups,” she said. “We hold active learning sessions to help them identify pain points in the organization where custom GPTs might be useful.”
One such learning session resulted in an HR manager developing a project management GPT for his team that walks them through how to build effective project plans.
Building HR Apps with Other Low-Code Platforms
In addition to custom AI assistants, some HR functions are using low-code and no-code development platforms to build their own HR software applications without the need for IT assistance.
Christopher Fernandez, corporate vice president of HR at Microsoft, is a pioneer in training HR professionals to be citizen developers. Fernandez has taught members of his HR team to build custom HR apps and agents using both Microsoft Power Apps and the company’s Co-Pilot Studio low-code development platform. HR pros at Microsoft have created a variety of apps and agents for processes such as career development, new-hire onboarding, and answering employees’ frequently asked HR questions.
To build the career development agent, for example, an HR team member had to find and extract authoritative data from both inside and outside of Microsoft.
“The information is tailored to specific professions here and is used by employees to determine how they can best develop their Microsoft careers through formal and informal training,” Fernandez said.
Fernandez believes domain expertise, not technical knowledge, is the key to building effective HR apps and agents.
“The vital skill is not a technologist-oriented skill; it’s deep domain expertise in HR workflows,” Fernandez explained. “The agents we’ve built have given my HR colleagues more reach around the world with their services.”
Fernandez said teaching HR staff to build agents using Microsoft’s low-code platforms is essentially a “paint-by-numbers” process. HR users answer a series of questions to guide them through creating agents that automate work that was previously manual.
“We teach our people to first have complete thoughts of what they’re trying to build and to ensure the tools are created within the framework of responsible AI,” Fernandez said. HR staff are also instructed to build apps to fit into the larger architecture of existing apps already created by their HR peers.
Testing and quality control are also important aspects to what these citizen developers in HR do.
“We want to ensure they’re connecting their agents to the most authoritative data sources we have,” Fernandez said. “Once that’s confirmed and tested, we move into deployment.”
Risks of Using Citizen Developers
The use of citizen developers to create HR apps or custom AI assistants doesn’t come without risks. Experts say nontechnical developers need clear guardrails and detailed training to ensure they don’t put sensitive company data at risk — and that they abide by regulations governing data privacy and retention.
“While programming languages aren’t used on these platforms, building apps with them still requires a certain degree of technical know-how,” Sheth said. “Without adequate governance and guidance, unchecked citizen developers can present challenges in terms of data security or complying with data privacy regulations like the General Data Protection Regulation. You can run into problems like citizen developers in HR storing sensitive data in noncompliant locations, for example.”
Courneen takes significant precautions to ensure the custom GPTs created in his HR function have the proper data security protections. For starters, he ensures everyone using the company’s enterprise ChatGPT account starts with the proper settings.
“That means going to the data tab and turning off the setting that says ‘improve the model for everyone’ — which is telling ChatGPT you’re not letting it keep your data,” Courneen said.
But he knows that’s only the first step in protecting MSI’s confidential data and any proprietary work processes.
“Information shared will still exist somewhere on the external vendor’s server,” he explained. “So, we’ve also trained our people — if they intend to upload any sensitive company information or work processes we feel give us a competitive advantage — that they shouldn’t use ChatGPT at all.”
Outside of Courneen’s benefits-related GPT, the other custom GPTs he built rely only on publicly available data from the internet and no confidential company information.
As HR leaders consider whether to create citizen developers on their own teams, those who’ve been down the path suggest setting aside any concerns over HR’s lack of technical skills or software development know-how.
“People just need to dive into these intuitive platforms and start building,” Pearson said. “I never for a minute thought I’d be building my own chatbots, but here I am creating a lot of useful tools over the past few years. The sky is the limit if you’re willing to jump in and start experimenting."
Dave Zielinski is a Minneapolis-based business journalist who covers the impact of emerging technologies on the workplace. He is a frequent contributor to SHRM publications.
Was this resource helpful?
Validate your HR expertise
Earning your SHRM-CP credential makes you a recognized expert and leader in the HR field.
Related Content
Learn how Marsh McLennan successfully boosts staff well-being with digital tools, improving productivity and work satisfaction for more than 20,000 employees.
The proliferation of artificial intelligence in the workplace, and the ensuing expected increase in productivity and efficiency, could help usher in the four-day workweek, some experts predict.
As artificial intelligence technology continues to develop, the demand for workers with the ability to work alongside and manage AI systems will increase. This means that workers who are not able to adapt and learn these new skills will be left behind in the job market.