How the workforce is being rebuilt from 140 systems into seven superagents. Featuring Josh Bersin.
Every vendor now sells AI agents, but most are selling a smarter chatbot. This guide walks Josh Bersin HR 2030 argument in full: the 140-system problem, the seven superagents that replace it, the context layer that lets them actually reason, and an 18-month roadmap for where to start on Monday.
Free download
Enter your details to download The Architecture of Agentic HR, featuring Josh Bersin. The complete blueprint, the seven superagents, and the 18-month roadmap.
Founder and CEO, The Josh Bersin Company
For more than two decades, Josh Bersin has been one of the most influential analysts in HR and workforce technology, advising many of the world largest enterprises on talent, learning, and the systems that run them. His HR 2030 research, the basis for this guide, reframes the category around agentic AI. When Bersin describes where HR technology is heading, the industry pays attention.
The case for reinvention
Bersin traces four eras of enterprise AI. A chatbot and an assistant retrieve and transact. A superagent reasons. Most vendors plateau at the second box.
Answers FAQs
Retrieves answers to common questions. Helpful, but it only ever responds.
Looks things up, fills forms
Finds information and completes transactions on request. You are still the one driving.
Takes multi-step action
Executes a sequence of steps across systems to complete a defined task.
Reasons, orchestrates, decides
Reasons over workforce context, coordinates specialist agents, and owns the outcome end to end.
A chatbot retrieves and an assistant transacts. A superagent reasons over your workforce context, coordinates a team of specialist agents, and owns an outcome end to end. The guide maps the four stages of autonomy so you can tell which one a vendor is really selling.
Bersin redraws all of HR on a clean sheet into seven outcome-owning superagents: Talent Acquisition and Mobility, Talent Development and Growth, Pay and Total Rewards, Employee Services and Support, Employee Experience and Workplace, Workforce Management, and HR Intelligence and Strategy.
The context layer is the rules and rails, the information that matters about your company: cultural values, business definitions, management practices, and guardrails. Without it an agent can retrieve and transact but cannot reason. Gloat Loomra is this layer in practice.
Modeled on a 1,000-person HR org, 30 to 40 percent of roles change, they do not vanish. Transactional functions compress hardest, while strategy, HR tech, and analytics hold or grow. The work moves up the stack.
Not with seven superagents on day one. The guide lays out an 18-month roadmap: stand up the context engine, ship one Tier-3 use case in the flow of work, then orchestrate multiple agents on real outcomes like redeployment and succession.
See it in your workforce
See Loomra and Galileo applied to your own roles, skills and systems. A working walkthrough of agentic HR on your data, with your team in the room.