IBM AskHR: How Agentic AI Handles 11.5 Million HR Queries a Year and Cut HR Costs by 40%
By Chris Weinmann, Founder, OVI
When IBM's HR team was fielding growing volumes of employee queries — questions about payroll, benefits enrollment, leave balances, and onboarding checklists — the math stopped working. Every routine inquiry that reached a human HR partner consumed time that could have gone toward workforce strategy, talent development, or organizational change. The company needed a way to handle volume without hiring proportionally more HR staff.
The answer was AskHR, an AI-powered HR service delivery platform that IBM deployed internally as its own "client zero." Through a six-year iterative approach and multiple platform upgrades, AskHR now runs on watsonx Orchestrate with full agentic AI capabilities — and the numbers tell the story.
The Scale: 11.5 Million Interactions in a Single Year
In 2024, AskHR processed 11.5 million employee interactions across IBM's global workforce (IBM case study; Codebridge). The platform automates more than 80 distinct HR tasks, from payroll queries and benefits enrollment to compliance checks and leave management (Supalabs; Codebridge).
The standout metric: a 94% self-service containment rate. Only 6% of interactions require escalation to a human HR partner (SHRM; Supalabs; Codebridge).
The Financial Impact: 40% Cost Reduction, $5M Annual Savings
The business case is concrete. Over four years, AskHR delivered a 40% reduction in HR operational costs (IBM case study; Codebridge). In absolute terms, that translates to $5 million in annual savings and 50,000 hours of HR team capacity freed up each year (IBM case study). Support tickets dropped 75% from baseline since the platform's inception (Codebridge).
For managers, the impact is even more immediate. HR transactions — promotions, team changes, approvals — now complete 75% faster, contributing to a 99% manager adoption rate (Supalabs; Codebridge). AskHR's HR automation was one contributor to IBM's broader $3.5 billion in productivity savings across the company in 2024 (IBM case study).
The Architecture: Two-Tier Support With Agentic AI
AskHR operates a two-tier support model. The AI layer handles routine, high-frequency requests — the kind that make up the bulk of any HR inbox. When a query involves a complex edge case, a policy exception, or a sensitive employee-relations matter, it escalates to a human HR advisor with full context attached.
The platform integrates with Workday, SAP, and Concur, pulling real-time data rather than relying on static knowledge bases (IBM case study). This integration layer is what enables AskHR to complete transactional tasks — not just answer questions.
In 2025, IBM upgraded AskHR with watsonx Orchestrate, adding agentic AI capabilities that allow the system to chain multi-step workflows autonomously (IBM announcement; Supalabs). Instead of responding to a single query, AskHR can now orchestrate sequences — for example, processing a parental-leave request by checking eligibility, calculating entitlements, updating payroll, and notifying the employee's manager in one automated flow.
What Changed for IBM's HR Team
IBM CEO Arvind Krishna has been direct about the workforce impact: AI agents automated tasks previously handled by hundreds of HR professionals focused on repetitive processing work (SHRM). But the story isn't one of simple headcount reduction. IBM redirected those resources toward software development, sales, and marketing — and the company's overall headcount actually increased (SHRM).
As Nichol Bradford, SHRM's executive in residence for AI, noted: "IBM was able to clean up a lot of the rote tasks people do. It doesn't mean that it's not a directional signal of change, but it is not the end of HR" (SHRM).
This pattern — AI absorbing transactional work while HR professionals move to strategic roles — is accelerating across the industry. According to SHRM's 2026 State of AI in HR report, 39% of organizations now have AI deployed in HR functions, with adoption concentrated in recruiting (27%), HR technology (21%), and learning and development (17%) (SHRM State of AI 2026).
What Other CHROs Can Replicate
IBM's six-year journey offers a replicable playbook — not because every company has IBM's budget, but because the principles scale down:
- Start with volume. Identify the HR queries that consume the most human hours. For most organizations, these are benefits questions, leave requests, and payroll inquiries.
- Measure containment, not just deployment. IBM tracks the percentage of interactions resolved without human involvement. This single metric separates useful AI from expensive chatbots.
- Keep humans in the loop for complex cases. The two-tier model is not a cost-cutting shortcut — it's a quality assurance mechanism. Edge cases handled poorly by AI erode employee trust faster than slow response times do.
- Integrate with systems of record. AskHR's ability to pull data from Workday, SAP, and Concur is what allows it to execute transactions, not just answer questions. Without integration, AI assistants remain FAQ bots.
- Iterate over years, not months. IBM refined AskHR across six years of continuous development. The 94% containment rate didn't arrive overnight — it compounded through iterative training, expanded task coverage, and the 2025 agentic upgrade.
The gap between companies experimenting with AI in HR and those achieving measurable results is often a gap in execution discipline. IBM's case demonstrates that sustained, well-integrated deployment — starting with routine tasks and expanding methodically — delivers compounding returns.
Frequently Asked Questions
What is IBM AskHR and how does it work?
IBM AskHR is an AI-powered HR service delivery platform that uses natural language processing to handle employee queries across 80+ HR tasks — including payroll, benefits, onboarding, compliance, and leave management. It operates a two-tier model: AI resolves routine requests autonomously, while complex edge cases escalate to human HR advisors.
What results has IBM achieved with AskHR?
IBM reports a 94% self-service containment rate, 40% reduction in HR operational costs over four years, 75% fewer support tickets, 50,000 hours saved annually, and $5 million in yearly savings. The platform processed 11.5 million employee interactions in 2024 alone.
How long did it take IBM to reach these results?
IBM's AskHR deployment followed a six-year iterative approach. Rather than a big-bang rollout, IBM expanded capabilities gradually — adding new HR tasks, integrating with enterprise systems like Workday, SAP, and Concur, and upgrading to agentic AI via watsonx Orchestrate in 2025.
Can mid-size companies replicate IBM's AI HR service delivery model?
Yes, though at a different scale. The core principles — starting with high-volume routine queries, measuring containment rates, maintaining human escalation paths, and expanding iteratively — apply regardless of company size. Several AI HR platforms now offer similar capabilities at price points accessible to mid-market organizations.
Does AI HR service delivery eliminate the need for human HR professionals?
No. IBM's model demonstrates the opposite: AI handles routine, repetitive inquiries so that human HR professionals can focus on strategic work like talent development, organizational design, and complex employee relations. IBM's overall headcount actually increased despite reducing routine HR processing roles.
Sources: IBM AskHR case study, IBM AI-first HR blog, IBM AI agents announcement, SHRM: IBM embraces AI, Supalabs IBM case study, SHRM State of AI in HR 2026, Codebridge: 10 AI HR case studies
What is IBM AskHR and how does it work?
IBM AskHR is an AI-powered HR service delivery platform that uses natural language processing to handle employee queries across 80+ HR tasks — including payroll, benefits, onboarding, compliance, and leave management. It operates a two-tier model: AI resolves routine requests autonomously, while complex edge cases escalate to human HR advisors.
What results has IBM achieved with AskHR?
IBM reports a 94% self-service containment rate, 40% reduction in HR operational costs over four years, 75% fewer support tickets, 50,000 hours saved annually, and $5 million in yearly savings. The platform processed 11.5 million employee interactions in 2024 alone.
How long did it take IBM to reach these results?
IBM's AskHR deployment followed a six-year iterative approach. Rather than a big-bang rollout, IBM expanded capabilities gradually — adding new HR tasks, integrating with enterprise systems like Workday, SAP, and Concur, and upgrading to agentic AI via watsonx Orchestrate in 2025.
Can mid-size companies replicate IBM's AI HR service delivery model?
Yes, though at a different scale. The core principles — starting with high-volume routine queries, measuring containment rates, maintaining human escalation paths, and expanding iteratively — apply regardless of company size. Several AI HR platforms now offer similar capabilities at price points accessible to mid-market organizations.
Does AI HR service delivery eliminate the need for human HR professionals?
No. IBM's model demonstrates the opposite: AI handles routine, repetitive inquiries so that human HR professionals can focus on strategic work like talent development, organizational design, and complex employee relations. IBM's overall headcount actually increased despite reducing routine HR processing roles.