Master Agentic HR in 4 tracks
58 articles across 4 tracks. From foundations to implementation. Start anywhere, but we recommend the path below.
Agentic HR
See all 50 terms →/a-jen-tik ach-ar/ · noun
A category of workforce technology where AI agents autonomously understand, reason, and act on workforce decisions in the flow of work - delivering intelligent, not just transactional, outcomes by drawing on cross-system context, business logic, and governed action.
Your learning tracks
Foundations of Agentic HR
The portal nobody visits: why HR tech has an adoption problem
Enterprise HR tech has an adoption ceiling: most organizations see 20-40% active usage of their HCM portals for…
From systems of record to systems of action
Recording is not acting. A system that stores compensation data is different from one that notices comp fell…
EssentialWhat agentic actually means and what it does not
The word agentic is everywhere. This article cuts through the noise with a clear definition, a capability spectrum,…
The copilot ceiling: why Q&A chatbots are not enough
Copilots deliver real value, but they hit a hard ceiling. This article explains the four structural constraints that…
Most citedThe session problem, the initiation problem, the learning problem
This is the most important article in the academy. It provides a deep framework for the three structural…
Autonomous action: what it means for an agent to do, not just answer
The difference between a chatbot and an agent is not intelligence. It is autonomy - the ability to…
Portal architecture vs. agent architecture
The architectural foundations of HCM platforms - portal delivery, siloed data, and human-initiated workflows - are structural barriers…
Business process frameworks were built for humans, not agents
The workflow engines inside HCM platforms were designed for human-driven, sequential processes. Agents need event-triggered, parallel, composable playbooks.
The single-system intelligence problem
Every major vendor is building AI assistants - but each one can only see the data inside its…
People context: beyond the employee record
Ring 1 of workforce context. The employee record captures job title, department, and compensation. Real people context captures…
Company context: org structure vs. how work actually happens
Ring 2 of workforce context. Every company operates differently. Policies, team structures, collaboration patterns, and unwritten rules shape…
Industry and world context: the rings no HCM has
Rings 3 and 4 of workforce context. Market benchmarks, skills supply and demand, regulatory intelligence, and network effects…
The business does not speak skills
HR teams have spent years building skills taxonomies, but the business still thinks in terms of work, outcomes,…
Taxonomic vs. inferential skills intelligence
Most enterprise skills platforms rely on taxonomic classification: humans or rules assign labels from a predefined list. Inferential…
From skills ontology to workforce intelligence
Organizations that stop at building a skills ontology mistake the raw material for the finished product. Skills data…
Consumption-based pricing and value-aligned economics
Traditional HR technology pricing charges per employee per year regardless of value delivered. Consumption-based pricing flips that model,…
Measuring agent outcomes: beyond adoption metrics
Most organizations measure HR technology success by adoption rates and login frequency. These vanity metrics reveal nothing about…
CapstoneBuilding the business case for your board
Board-level approval for agentic HR requires more than a vendor comparison. It demands a structured business case that…
Architecture and Technology
Loomra anatomy: five layers explained
The Intelligence Engine is not a single algorithm. It is five discrete layers, each with a clear contract,…
How Loomra ingests your workforce data: ingestion, harmonization, enrichment
The Context Engine is the first layer of the Intelligence Engine. It connects to source systems, harmonizes disparate…
Knowledge Graph, Retrieval, and workforce-specific embeddings
Generic search treats workforce data as flat text. A knowledge graph with 2.4 million nodes and 18.7 million…
The Agentic Layer: agent runtime, goal engine, action gateway
The intelligence layer knows. The agentic layer acts. Three components divide the work of execution: the Agent Runtime…
Coexistence with Workday: APIs, security groups, write-back
Workday is the system of record for millions of workers. Gloat does not replace it. Gloat coexists with…
Coexistence with SAP SuccessFactors: OData, RBP, Compound Employee
SAP SuccessFactors has its own API idioms, its own security model, and its own AI assistant. Gloat coexists…
Living in Teams and Slack: why agents belong in the flow of work
An agent that lives behind a portal login is an agent that gets ignored. The highest-performing workforce agents…
Why generic embeddings fail for workforce decisions
A generic embedding model thinks a Java developer and a JavaScript developer are nearly identical because the words…
Matching engine, skills clustering, and trajectory models
Traditional HR matching is keyword overlap. Modern matching engines score across skills, experience, trajectory, and organizational context to…
RAG, HyDE, and semantic retrieval for HR context
An LLM without retrieval hallucinates confidently. RAG grounds generation in your actual data. HyDE improves retrieval by generating…
Governed autonomy: the spectrum from suggestion to action
The autonomy spectrum runs from passive suggestions through guided workflows to fully autonomous actions. Each level has different…
EU AI Act readiness and bias auditing for HR agents
The EU AI Act classifies employment-related AI as high-risk. Compliance requires risk management, data governance, transparency, human oversight,…
Explainability: how agents show their reasoning
Agentic HR systems make consequential decisions about people's careers. Explainability turns black-box outputs into auditable, understandable reasoning that…
Six dimensions for evaluating HR agent architecture
The market for agentic HR technology is crowded and confusing. This article introduces a six-dimension evaluation framework that…
Build vs. buy: when to use horizontal AI platforms for HR
Microsoft Copilot Studio, Azure AI, and Amazon Bedrock are powerful platforms. But using them to build HR-specific agent…
The vendor landscape: Joule, Sana, Copilot, and beyond
The agentic HR vendor landscape is crowded and fast-moving. This article defines five vendor archetypes, examines the architectural…
Agent Use Cases in Practice
Why succession planning breaks the moment the spreadsheet closes
Succession planning typically peaks during a single annual review and decays immediately after. This article explores why static…
Critical role identification: beyond senior equals critical
Most organizations equate critical roles with senior titles. But a principal engineer who holds institutional knowledge about a…
Readiness as a vector, not a label
Traditional readiness labels flatten a complex, multi-dimensional assessment into a single bucket. This article explains how agentic systems…
The internal mobility problem: invisible talent, invisible roles
Every year, talented employees resign to take roles at other companies that their own organization had open internally.…
Career pathing with trajectory intelligence
Career pathing used to mean browsing a job board. With trajectory intelligence, an agent maps personalized paths based…
Internal-first hiring: surfacing talent before going external
Most organizations say they prioritize internal mobility. Few have the infrastructure to actually do it at the moment…
Flight risk is not a score - it is a pattern
A single flight risk score tells you almost nothing actionable. Behavioral pattern detection across multiple signals tells you…
Proactive retention interventions: what agents can trigger
Knowing someone might leave is only valuable if you can act fast enough to change the outcome. Agents…
The manager as agent consumer: real-time team intelligence
Managers make the most consequential talent decisions in most organizations. They also have the least access to workforce…
Workforce scenario planning: modeling the impact of AI on roles
A CHRO asks what happens if the company automates 30% of customer service. The agent models three scenarios,…
Skills gap detection and learning matching
60% of a data team lacks MLOps skills critical to next year strategy. The agent maps the gap…
Redeployment over layoffs: the ethical and economic case
A business unit restructures and 120 roles are eliminated. Instead of severance packages, the agent matches affected employees…
Intelligence sharing vs. data sharing
Data sharing moves rows between systems. Intelligence sharing moves meaning. The Shared Context Engine is what makes the…
The compound effect: why the platform gets smarter over time
Outcome tracking, feedback loops, and network effects create a flywheel where each interaction makes the next one better.
Agent Studio and Playbooks: configuring what your agent can do
Agent Studio lets HR teams configure what agents can do, when they act, and what guardrails apply, without…
Glossary and Reference
A-Z glossary of agentic HR
The definitive reference for agentic HR terminology. Forty-five terms defined, explained, and contextualized for HR leaders, IT partners,…
Terms the industry gets wrong
The agentic HR space has a language problem. Terms get stretched, conflated, and repurposed until they lose meaning.…
The jargon decoder: translating vendor marketing into reality
Vendor marketing in the agentic HR space has outpaced vendor engineering by roughly 18 months. Here is a…
Agentic HR vs. SAP Joule
SAP Joule is a copilot embedded in SuccessFactors modules. It assists within the application. Agentic HR operates across…
Agentic HR vs. Workday Sana
Workday Illuminate brings AI capabilities to the Workday ecosystem. The architecture is portal-centric, BPF-constrained, and bounded by Skills…
Agentic HR vs. Microsoft Copilot for HR
Microsoft Copilot is an infrastructure layer that adds generative AI to the Microsoft 365 surface area. Agentic HR…
The agentic HR readiness assessment
Deploying agentic HR is not purely a technology decision. It requires data maturity, process readiness, governance structures, and…
TemplateRFP template: questions to ask every vendor
Most HR technology RFPs recycle generic procurement language that fails to surface the differences between agentic platforms. This…
Implementation roadmap: from first agent to platform standardization
The most common implementation failure is trying to deploy everything at once. This roadmap breaks agentic HR adoption…
Ready to see Agentic HR in action?
The academy teaches the concepts. A demo shows the reality.
Get a demo