AMD's 30,000-Employee Workforce Has an AI HR Agent Now — And It Cut Resolution Time by 80%
AMD's 30,000-Employee Workforce Has an AI HR Agent Now — And It Cut Resolution Time by 80%
When a company that designs the chips powering AI deploys agentic AI inside its own HR function, the industry pays attention. AMD, the $200-billion semiconductor maker, rolled out Kore.ai's agentic AI platform across its 30,000-person global workforce — and the Kore.ai-reported results are hard to ignore: 80% faster HR resolution times, 50% of employee queries handled entirely through self-service, and a 70% jump in employee satisfaction scores.
AMD is, in the most literal sense, eating its own cooking.
The Problem: A Lean HR Team Stretched Across Time Zones
AMD operates across dozens of countries, with engineering hubs, manufacturing sites, and sales offices spanning North America, Europe, and Asia-Pacific. Like most global enterprises, its HR helpdesk fielded a constant stream of repetitive queries — benefits questions, policy lookups, leave balances, payroll clarifications — that consumed bandwidth a lean HR team could not afford to lose.
The challenge was structural, not personnel. HR professionals were spending the bulk of their time on high-volume, low-complexity requests, leaving insufficient capacity for the strategic work — workforce planning, talent development, organizational design — that a chipmaker in the middle of an AI hardware boom desperately needed.
As Veeru Teki, Senior Architect of IT Business Systems at AMD, detailed in a November 2025 webinar, the company needed a system that could handle routine queries autonomously while preserving human oversight for sensitive matters (Kore.ai AI Activations Webinar, Nov 19, 2025).
The Solution: Kore.ai's Agentic AI for HR
AMD deployed Kore.ai's agentic AI platform with capabilities that go beyond a standard chatbot. The system provides:
- Unified HR information access: Employees get instant, accurate answers to policy and benefits questions across regions, without navigating multiple portals or waiting for a ticket response.
- Self-service transactions: Common actions — updating personal details, checking leave balances, submitting routine requests — are handled end-to-end by the AI agent without human intervention.
- Role- and region-specific responses: The platform tailors answers based on the employee's location, role, and applicable policies, reducing the confusion that comes with a globally distributed workforce.
- Escalation to HR professionals: For sensitive issues — performance concerns, accommodations, complex benefits disputes — the agent routes the conversation to a human HR professional with full context attached.
That last point matters. AMD did not deploy AI to replace its HR team. It deployed AI to redirect its HR team's attention to the work that requires human judgment.
The Results: Kore.ai-Reported Metrics
According to Kore.ai's published case study, AMD's deployment produced three headline outcomes (Kore.ai Blog):
| Metric |
Result |
| HR resolution time |
80% reduction (Kore.ai-reported) |
| Self-service adoption |
50% of queries resolved without human intervention (Kore.ai-reported) |
| Employee satisfaction |
70% increase (Kore.ai-reported) |
These are vendor-reported figures, not independently audited. They are directionally informative and consistent with outcomes reported by other large-scale agentic HR deployments, but HR leaders should treat them as indicative rather than definitive.
Robert Gama, SVP & Chief Human Resources Officer at AMD, framed the initiative in strategic terms (Kore.ai Blog):
"As a global leader in AI, we saw a clear opportunity to bring leadership in our own workplace. Our work with Kore.ai shows what's possible when you use AI not to replace people, but to enhance how they work, connect, and lead."
Why the "Eating Its Own Cooking" Angle Matters
AMD is not a consulting firm running a pilot. It is one of the three companies — alongside NVIDIA and Intel — whose hardware makes generative and agentic AI possible at scale. When AMD's CHRO deploys agentic AI internally and reports meaningful results, it sends a specific signal to the market: this technology is mature enough for the companies building it to bet on it internally.
That credibility gap — between vendors who sell AI and organizations that actually use it — remains one of the biggest obstacles to enterprise AI adoption. AMD's deployment narrows it.
What Humans Still Handle
AMD's model preserves clear boundaries. The AI agent handles the routine; humans handle the rest. Escalation triggers include:
- Sensitive employee relations: Harassment reports, grievance processes, and disciplinary matters stay with trained HR professionals (Kore.ai Blog).
- Complex benefits disputes: Cases requiring interpretation of plan documents or cross-jurisdictional coordination are routed to specialists.
- Accommodation requests: Accommodation requests — including ADA and international equivalents — are escalated to HR professionals for individual assessment (Codebridge — AI in HR Case Studies).
This is not a footnote. It is the architecture. The system was designed with escalation as a first-class feature, not an afterthought (Codebridge — AI in HR Case Studies).
Industry Context
AMD's deployment lands at a moment when agentic AI in HR is moving from experimental to operational. According to SHRM's State of AI in HR 2026 report, 87% of CHROs forecast greater AI adoption in HR processes this year (SHRM, 2026). The question for most enterprises is no longer whether to deploy AI in HR, but how to deploy it without losing the human element that makes HR effective.
AMD's answer — agentic AI for the routine, human professionals for the complex — is emerging as the operational consensus. The results, while vendor-reported, suggest that the model works at scale.
Source Attribution: