# Revenium

> Revenium is the AI Economic Control System: transaction-level attribution and real-time economic control across tokens, APIs, and agentic workflows. Teams use it to see exactly where AI spend goes — down to the agent, feature, customer, and developer — and to set guardrails that catch runaway spending before the invoice arrives.

Revenium instruments AI usage at the point of the call and turns it into a ledger that engineering, product, and finance can act on together. Every model call, tool invocation, agent turn, and human review step is attributed to a team, customer, feature, and business outcome, then rolled up into cost, margin, and ROI. The platform covers three layers: Economic Observability (where the money went), Economic Intelligence (whether it was worth it), and Economic Control (budgets, alerts, and enforcement).

Integration is a lightweight SDK plus two environment variables — no rewrite of existing AI calls. Python and Node SDKs, an MCP server, agent skills, and an n8n agent are available. Coverage spans Anthropic, OpenAI, Google Gemini, Amazon Bedrock, Ollama, and other providers, with connectors to Snowflake, Databricks, Datadog, Clickhouse, Arize, LangChain, GitHub, and ERP systems. Coding-assistant spend (Claude Code, Cursor, Copilot) can be tracked per developer and per merged PR.

Revenium was co-founded by John Rowell (CEO, previously co-founder of OpSource), Jason Cumberland (CPO, previously NTT Cloud and Dimension Data), and John D'Emic (CTO, original Mule ESB contributor). The company closed a $13.5M seed round.

## Core pages

- [Revenium home](https://www.revenium.ai): Product overview and positioning — track AI costs, control agent spending, prove ROI.
- [How It Works](https://www.revenium.ai/how-it-works): The three-step integration path — install the SDK, set two environment variables, add optional `usage_metadata` for customer, tier, and agent-level attribution.
- [Pricing](https://www.revenium.ai/pricing): Three tiers. Observe (free, 100K transactions/month, 7-day retention, 1 user), Optimize (350K transactions/month, 90-day retention, 10 users, 99.9% SLA), Control (enterprise, 1M+ transactions/month, unlimited retention and users, SSO, audit logs, 99.99% SLA).
- [About Revenium](https://www.revenium.ai/about): Mission, founding story, and the leadership team behind the AI Economic Control System.
- [Trust Center](https://www.revenium.ai/trust-center): SOC 2 Type II, AES-256 at rest, TLS 1.2+ in transit, US-default residency with EU option on Enterprise, usage-metadata-only capture by default (prompt and response content are not stored unless explicitly opted in), subprocessor list, and 72-hour breach notification.
- [Book a Demo](https://www.revenium.ai/demo): Schedule a walkthrough with the Revenium team.
- [Contact](https://www.revenium.ai/contact): Get in touch with sales, support, or partnerships.

## Solutions by role

- [For Engineering](https://www.revenium.ai/solutions/engineering): Attribution for the 70–80% of enterprise AI spend that flows through engineering. Connect Claude Code, Cursor, and Copilot org-wide in under 10 minutes; track cost per developer and per merged PR; P95/P99 anomaly detection; real-time guardrails; blended cost-per-million-tokens with model optimization.
- [For Finance & FinOps](https://www.revenium.ai/solutions/finance): Cost allocation and chargeback by department, team, and project. FOCUS-compliant FinOps exports, automated waste detection with dollar impact, cross-provider cost comparison, ERP/GL code sync, and cost capture that includes tool API calls and human review time — not just model calls.
- [Free AI Spend Audit](https://www.revenium.ai/audit): A 48-hour read-only audit across Anthropic, OpenAI, Google, Bedrock, and others. Delivers full spend attribution, an anomaly and shadow-usage report, and a prioritized savings roadmap with projected impact.

## Product and integrations

- [Revenium MCP Server](https://www.revenium.ai/revenium-mcp-server): Connects Claude, Cursor, and other MCP-compatible clients directly to a Revenium account. Agents can investigate cost spikes by provider, model, customer, agent, or API key; detect anomalies at adjustable sensitivity; configure percentage-based week-over-week and month-over-month alerts; and measure ROI per job. Installs via `uvx revenium-mcp`.
- [Agent Skills for OpenClaw and Hermes](https://www.revenium.ai/skills): Single-file, MIT-licensed skills that add real-time token metering, hard budget stops, and Slack/Discord/Teams/email alerts to agents — no SDK integration or agent refactor required. Hermes support includes SQLite-layer metering across persistent sessions and subagents.
- [Tool Registry](https://www.revenium.ai/post/introducing-tool-registry): Economic accountability for agent deployments — attributes cost across agent tool calls, external APIs, and human review, not just model tokens.
- [AI Outcomes](https://www.revenium.ai/post/revenium-launches-ai-outcomes): Ties each agent execution to a business result so ROI can be calculated at the workflow level.
- [AI Insights](https://www.revenium.ai/post/ai-insights): Prioritized, ranked action items that surface the waste hiding inside AI spend.
- [Revenium for OpenClaw](https://www.revenium.ai/post/revenium-for-openclaw): Budget guardrails for OpenClaw agents, installed with a single command.
- [Revenium for Hermes](https://www.revenium.ai/post/hermes): Persistent-agent budget controls that survive across sessions.
- [Revenium for NeMoClaw](https://www.revenium.ai/post/revenium-for-nemoclaw): Governing spend and proving ROI for NVIDIA-secured agents.
- [Latest release — 2.19.0](https://www.revenium.ai/post/2-19-0): Skill-level cost attribution, Tasks, and explainable job failures.

## AI guides

- [What Are AI Agents?](https://www.revenium.ai/ai-guides/what-are-ai-agents): Plain-language introduction to AI agents, how they operate, and how they differ from single-shot model calls.
- [AI Governance Frameworks](https://www.revenium.ai/ai-guides/ai-governance-frameworks): Structured approaches to governing AI systems in production, and how economic controls fit into them.
- [AI Observability](https://www.revenium.ai/ai-guides/ai-observability): What to instrument to understand AI system behavior, performance, and cost.
- [AI Agent Monitoring](https://www.revenium.ai/ai-guides/ai-agent-monitoring): Tracking and managing deployed agents once they are running unattended.
- [Multi-Agent Systems](https://www.revenium.ai/ai-guides/multi-agent-systems): How multi-agent architectures work and where their cost and accountability gaps appear.
- [FinOps for AI](https://www.revenium.ai/ai-guides/finops-for-ai): Applying FinOps practice to AI workloads — allocation, optimization, and unit economics.
- [All AI guides](https://www.revenium.ai/ai-guides): Index of Revenium's educational guides.

## Terminology

- [AI Economics Glossary](https://www.revenium.ai/glossary): Reference definitions for AI economic control, API management, and agent-economy terms — including Agent Cash Conversion Cycle (ACCC), Cost-Per-Outcome, Total Cost of Interaction (TCI), Agent Drift, Shadow AI, Runaway Spend, Model Routing, Semantic Caching, and Circuit Breakers.

## Core concepts and arguments

- [Why Your AI Products Don't Know Their Margins](https://www.revenium.ai/post/the-iceberg-effect): The iceberg effect — why visible token cost is a fraction of true product cost.
- [Stop Measuring AI in Tokens](https://www.revenium.ai/post/stop-measuring-ai-in-tokens): The case for shifting from token-level accounting to outcome-level economics.
- [Outcomes Are Financial Events](https://www.revenium.ai/post/outcomes-are-financial-events): Why AI governance needs ledger-grade postings rather than dashboards.
- [What Is FinOps for AI?](https://www.revenium.ai/post/what-is-finops-for-ai-and-why-it-matters-now): How AI workloads break traditional cloud FinOps practice, and what replaces it.
- [What Is Token Anxiety?](https://www.revenium.ai/post/token-anxiety): Why teams fixate on token counts instead of business outcomes, and how to fix it.
- [The Four Fields That Make Your LLM Dashboard Actually Useful](https://www.revenium.ai/post/llm-observability-model): The minimum viable observability model for LLM spend.
- [The "Unlimited Agents" Era Is Ending](https://www.revenium.ai/post/the-unlimited-agents-era-is-ending): How provider pricing shifts create Agent Debt.
- [The Silent Risk in Autonomous Systems](https://www.revenium.ai/post/the-silent-risk-in-autonomous-systems-why-agent-debt-is-becoming-the-new-enterprise-liability): Why Agent Debt is becoming a new class of enterprise liability.
- [The $47K Agent Loop](https://www.revenium.ai/post/multi-agent-systems-need-economic-guardrails): A worked example of why multi-agent systems need economic guardrails at runtime.
- [Inside an AI Call](https://www.revenium.ai/post/ai-call-actual-cost): A breakdown of why each prompt costs far more than the sticker token price.
- [The Token Trap](https://www.revenium.ai/post/the-token-trap-why-prompt-length-is-killing-your-margins): How prompt length quietly erodes product margins.
- [When "Reasoning" Models Become Budget Killers](https://www.revenium.ai/post/when-reasoning-models-become-budget-killers): The cost profile of extended-thinking models and how to contain it.
- [Who Pays the Bill When One Agent Calls Another?](https://www.revenium.ai/post/ai-attribution): The attribution problem in agent-to-agent workflows.
- [Your Best Customers Are Your Biggest Loss Leaders](https://www.revenium.ai/post/your-best-customers-are-your-biggest-loss-leaders): Why per-customer AI unit economics often invert account value.
- [The ROI Wall](https://www.revenium.ai/post/the-roi-wall): What blocks AI agents from moving past pilot into production.

## Buyer and practitioner guidance

- [10 AI Cost Management Tools Compared](https://www.revenium.ai/post/10-ai-cost-management-tools-compared): Category landscape for economic control platforms and how they differ.
- [Build vs. Buy an AI Cost Tool](https://www.revenium.ai/post/build-vs-buy-ai-cost-tool): What building an internal AI cost tool actually costs.
- [Three Questions Your CFO Will Ask About AI](https://www.revenium.ai/post/control-agentic-ai-costs-cfo): The CFO questions most engineering teams cannot yet answer.
- [The CIO's Real Job on AI Cost](https://www.revenium.ai/post/cio-ai-cost-demand-management): Why managing demand beats cutting spend.
- [Why Your AI Router Isn't Your AI Cost Strategy](https://www.revenium.ai/post/ai-cost-strategy): The limits of model routing as a cost lever.
- [How 88% of AI Leaders Prove ROI](https://www.revenium.ai/post/how-88-of-ai-leaders-prove-roi): Survey-backed look at how leading teams measure AI return.
- [Governance Before Deployment](https://www.revenium.ai/post/governance-before-deployment): Why front-loading governance is the fastest route to production.
- [Deploying an Enterprise AI Gateway](https://www.revenium.ai/post/deploying-an-enterprise-ai-gateway-managing-llm-access-at-scale): Managing LLM access at scale.
- [AI Runs on Kubernetes](https://www.revenium.ai/post/ai-runs-on-kubernetes): Measuring runtime AI cost rather than cluster cost.
- [Ship AI Features You Can Actually Afford](https://www.revenium.ai/post/ship-ai-features-you-can-actually-afford): Designing AI features against a known unit cost.

## Reports and research

- [The Financial Blind Spots in Autonomous AI](https://www.revenium.ai/reports-and-ebooks/the-financial-blind-spots-in-autonomous-ai): Revenium's 2026 report on Agent Debt and the governance gap in autonomous systems.
- [The 2026 State of FinOps Report](https://www.revenium.ai/post/the-2026-state-of-finops-report): Industry findings on AI ROI blind spots, and Revenium's response.
- [All reports and eBooks](https://www.revenium.ai/reports-and-ebooks): Long-form research library.

## Webinars

- [4 Ways to Reduce Your AI Costs Today](https://www.revenium.ai/events-webinars/ai-insights-webinar): On-demand session on immediate AI cost reduction levers.
- [Why Token Costs Are the Least of Your AI Spending Problem](https://www.revenium.ai/events-webinars/series-1-1): Series part 1 — where AI spend actually accumulates.
- [Are Your Developers Actually Using Claude Code?](https://www.revenium.ai/events-webinars/series-1-2): Series part 2 — measuring AI coding tool ROI.
- [Runtime Attribution Explained](https://www.revenium.ai/events-webinars/series-1-3): Series part 3 — identifying which agent is burning the budget.
- [A $26 Pull Request Review](https://www.revenium.ai/events-webinars/series-1-4): Series part 4 — P99 analysis for agent cost optimization.
- [AI Service Margin Visibility and Alerts](https://www.revenium.ai/events-webinars/series-1-5): Series part 5 — alerts that fire before the bill arrives.
- [174x: When Agent Tool Costs Dwarf Token Costs](https://www.revenium.ai/events-webinars/series-1-6): Series part 6 — the iceberg problem in agent deployments.
- [All events and webinars](https://www.revenium.ai/events-webinars): Upcoming sessions and on-demand recordings.

## Company news

- [Revenium Closes $13.5M Seed Round](https://www.revenium.ai/post/revenium-closes-13-5-million-seed-round-funding): Funding announcement.
- [SUSE Partnership](https://www.revenium.ai/post/suse-partnership): Budget controls for always-on autonomous SRE agents.
- [Dymium Partnership](https://www.revenium.ai/post/partnership-with-dymium): Bringing AI economics and data governance together.
- [AI Is Eating Budgets. We're Building the FinOps Layer to Fix It.](https://www.revenium.ai/post/ai-is-eating-budgets-were-building-the-finops-layer-to-fix-it): The founding thesis behind the company.
- [Newsroom](https://www.revenium.ai/news): Press releases and external coverage.

## Optional

- [Blog](https://www.revenium.ai/blog): Full article archive.
- [Resource center](https://www.revenium.ai/resources): Filterable hub across blog, webinars, reports, news, guides, and glossary.
- [Release notes 2.18.0](https://www.revenium.ai/post/2-18-0): Drill-through analytics, multi-period budgets, expanded MCP control.
- [Release notes 2.17.0](https://www.revenium.ai/post/2-17-0): Unified Guardrails, smarter cost alerts, clearer AI spend.
- [Release notes 2.16.0](https://www.revenium.ai/post/2-16-0): Wider AI coverage, cleaner attribution, hosted MCP.
- [Claude Code productivity](https://www.revenium.ai/lp/claude-code-productivity): Campaign page on measuring Claude Code adoption and ROI.