Revenue Contract Agent is in early access.
Meridian

Resources

Guides for teams putting agents to work

Meridian’s resources are 8 guides on governed AI agents for HR and finance: explainers, governance frameworks, finance operations, and buying advice. Every guide is signed, dated, and revised when the facts change.

All guides

FeaturedBuying and ROI

Measuring the ROI of AI agents: hours, dollars, and risk

Formulas and four worked examples for AI agent ROI in HR and finance, using modeled design-partner outcomes: hours recovered, direct dollars, risk value, what to exclude, and a 90-day measurement plan.

PR

Priya RamanHead of Finance Solutions

6 min read

In this guide

  1. 01AI agent ROI = (labor value + direct dollar value − agent cost) ÷ agent cost, with hours, dollars, and risk reported in separate columns rather than blended.
  2. 02Hours recovered per month = volume × automation rate × (baseline minutes − review minutes) ÷ 60; the hours count as savings only when redeployed, not backfilled, or absorbing growth.
  3. 03A 6,000-employee organization modeled at 4,800 HR cases a month, 70% automation, and $42 per hour recovers about 896 hours and $37,600 per month against a $2,499 Growth subscription.
Read the guide

Latest guides

Newest first. Each guide opens with a definition, states one claim per sentence, and ends with the questions buyers ask.

Buying and ROI

The 2026 buyer's guide to AI agents for HR and finance

A 12-question vendor checklist, the red flags that predict failed deployments, a five-step pilot design, seven success metrics, and a 14-week timeline from first call to scale decision.

Nadia Kowalski6 min read

Explainers

Model Context Protocol for enterprise agents: what CFOs and CIOs need to know

MCP explained without jargon: how agents discover and call tools, where tool permissions must be enforced, why the gateway pattern matters, and how OpenTelemetry makes agent runs observable.

Owen Gallagher6 min read

Finance operations

Closing the books 3 days faster with AI agents

How the Close, Controls, Audit, and Planning Agents remove three business days from the month-end close: orchestration, reconciliations, evidence, controls, and a before-and-after close calendar.

Priya Raman6 min read

Governance

How to govern AI agents in HR: a practical framework

A seven-part framework for governing AI agents in HR: scope statements, least-privilege data permissions, approval tiers, audit trail, model choice, employee communication, and works council and GDPR alignment.

Marcus Feld7 min read

Buying and ROI

Consumption pricing for AI agents, explained

How credit-based pricing works for AI agents: what counts as a billable action, Meridian's published credit rates, three worked examples, budgeting, and guardrails against overage.

Aisha Bello6 min read

Explainers

AI agents vs. chatbots in HR and finance: the difference that matters

Chatbots answer; agents complete work. The four differences that matter in HR and finance are scope, actions, approvals, and measurability, with a side-by-side comparison table.

Elena Marchetti6 min read

Explainers

What is an agent system of record?

A plain definition of the agent system of record, the 12 field groups it tracks, why it borrows from HR's employee registry, and nine criteria for evaluating one.

Tomas Lindqvist6 min read

Glossary

Terms, defined in one sentence

34 plain definitions of the vocabulary behind governed agents in HR and finance. Each entry is one citable sentence, with the long version a click away.

All 34 terms
Agent system of record
An agent system of record is the authoritative registry of every AI agent in an organization, recording owner, permissions, data access, and compliance status.
Governed agent
A governed agent is an AI agent confined to one workflow, permitted data, and logged actions, with human approval required for consequential decisions.
Human-in-the-loop
Human-in-the-loop is a design pattern in which an automated system pauses at defined points for a person to review, approve, edit, or reject before it proceeds.
Model Context Protocol (MCP)
The Model Context Protocol is an open standard that lets AI models and agents discover and call external tools and data sources through a common interface.
Month-end close
Month-end close is the accounting process of finalizing a period's transactions, reconciliations, accruals, and reviews so financial statements can be produced.
Consumption pricing (credits)
Consumption pricing charges for software by the work it performs, measured in units such as credits, rather than by the number of users who have access.

Get started

Put the first agent to work this quarter.

Start with one workflow, one approver, and one number to move. Most design partners were live in five weeks.

Book a demoView pricing