AskHR
Abstract illustration of two profile silhouettes, speech bubbles, pens, and papers
From legacy to leading edge

Our HR transformation journey began almost 10 years ago. At the time, we didn’t know how broadly it would reshape the organization. There was no grand plan—just a response to the challenges of the moment.

Our budget was shrinking even as our work grew more complex. At the same time, consumer technology outside IBM was becoming faster, more intuitive and more personal—raising expectations that inevitably followed employees into the workplace.

By 2016, those forces reached an inflection point. We realized it was time to try something bold. Traditional HR systems and operating models could no longer keep pace with the scale and speed of change underway. 

The moment called for a move that would reimagine how we worked and redefine what it meant to be an HR professional at IBM.   

Enter conversational AI chatbots.

Using IBM’s AI assistant builder, we started building chatbots to answer baseline HR questions about programs and benefits. Many worked well. Before long, 25 bots were in use.

There was just one problem. 

The more bots we added, the harder it became to get clear answers. When questions fell outside a chatbot’s scope, the user had to search elsewhere, leading to a fragmented employee experience. To restore coherence, we brought the bots together under a single interface: AskHR.

Shortly after AskHR launched, we informed all 21,000 first-line managers worldwide that dedicated HR business partner support would no longer be available. The global HR email address and phone number were also shut down. Managers could only access support through AskHR.  

It was a bold move. We believed in the technology and committed to full-scale adoption, recognizing the team would need time to adapt.

Over the following year, we monitored employee feedback closely. Our internal Customer Satisfaction Score (CSAT), a metric ranging from −100 to +100, fell to −35.

Person sitting on a magenta bench alcove using a laptop, with a notebook and water bottle
From adjustment to adoption

The early feedback made one thing clear: the organization was still adjusting to a new way of getting support. Over time, usage told a more complete story. 

Simply by using AskHR in their daily work, employees and managers helped refine it. Gradually, IBMers began to experience the value of an always-on, 24x7 resource. We expanded its capabilities accordingly, adding support for transactions.  

Suddenly, AskHR could do more than answer questions. It could take action, so managers could request an employee transfer or generate an employment verification letter, and AskHR would complete the transaction directly in the backend system.  

Across nearly 90,000 employee transfers each year, time and productivity gains added up quickly.

As results improved, we expanded the model. Between 2021 and 2024, AskHR—working in concert with our HR partners—became the single entry point for first-line managers and 85% of upline managers. 

And that -35 CSAT from the early days of AskHR?  It climbed to +74.  

Why? 

First, automation enabled immediate support for repeatable tasks.   

Second, refinements to our human support model accelerated how quickly managers could reach a live expert. 

Third, we introduced generative AI (gen AI) capabilities, increasing AskHR’s value to users.  

Two people seated at a round table talking, with cups and a phone
Find a small pain point. Automate that, make it great and then scale.

Nickle LaMoreaux

Senior Vice President and Chief Human Resources Officer

One of those capabilities was multiagent orchestration. In 2025, AskHR became the digital front door for all HR transactions, able to coordinate multiple autonomous agents to execute complex tasks together. 

That same year, AskHR handled more than 16 million individual user messages. That’s the equivalent of about 39,000 a day. Or roughly 51 interactions per IBMer per year.

The lessons we carried forward

We made some bold, even drastic, decisions on our journey. If we were to do it all again, we would approach some things differently. For starters, we would pilot and iterate. You don’t need to do everything at once. Think about the areas that make sense to focus on. Start small and then expand.  

We also learned that culture overrides everything. The tone leaders set around adoption speaks volumes. Empowering employees through skill-building and responsive feedback drives momentum. After all, without culture, it’s just technology.  

With another domain transformation just around the corner, IBM was ready to put those lessons to work. 

Explore the transformation

IBM’s move into AI-powered HR wasn’t linear. Experience made one thing clear: progress outperforms perfection. Redesigned for people, not hype, AskHR turned a legacy domain into a function built to scale.

Abstract illustration of two profile silhouettes, speech bubbles, pens, and papers

IT support was burdened by redundant processes. AskIT, an AI assistant that learns and resolves issues, turned the tides. IBM’s IT function became a self-service hub, with AskIT resolving 86% of routine issues.

Abstract illustration of two computer monitors connected by a blue bridge

IBM sellers were bogged down with system overload. Juggling multiple siloed sales tools created administrative sprawl. Then came AskSales, an AI lifeline co-developed internally to slash waste, automate basic tasks and turn a labyrinth into a fast lane to seller success.  

Open filing cabinet with hanging folders beneath a round clock

A bold, future-focused experiment is unfolding in the CIO sandbox. The ultimate vision is AskIBM, one AI-powered gateway to all things IBM. The beta is already showing promise, bringing coherence to complexity and continuing to redefine the future of our enterprise.

Illustration of a capsule descending by red parachute
Footnotes

© Copyright IBM Corporation 2026. IBM, the IBM logo and watsonx are trademarks of IBM Corp., registered in many jurisdictions worldwide.

Examples presented as illustrative only. Actual results will vary based on client configurations and conditions and, therefore, generally expected results cannot be provided.

Statements regarding IBM’s future direction and intent are subject to change or withdrawal without notice and represent goals and objectives only.