Documentation

The Zero-Access pack: menu paths to export journal templates, trial balances, AP, chart of accounts and budgets from seven systems. Nothing connected, nothing installed.

Connect QuickBooks Online, map cost pools to classes, and post locked AI chargeback months as balanced journal entries. Written for controllers: what Meridian can and cannot touch in your books, and why each setup step matters.

Connect Business Central, set up the MERIDIAN journal batch and permission set, and post locked AI chargeback months as general journal entries tagged with a Department dimension. Written for controllers and the admins who carry out the setup.

Send LiteLLM Generic API callback logs to Meridian gateway ingest, attribute calls by virtual key in the gateway key registry, and verify usage in the ledger.

Send Kong AI Gateway audit logs to Meridian with http-log, map Kong consumers to agents in the gateway key registry, and verify usage in the ledger.

Forward Portkey Logs Export records to Meridian gateway ingest on a schedule, attribute them by agent id metadata, and verify usage in the ledger.

Send DCGM GPU metrics from Prometheus remote write or the OpenTelemetry Collector to Meridian, map pods to endpoints, and split cost pools by GPU time.

Vendor-neutral discovery and current agent-spend metrics for any monitoring platform.

API reference · Changelog

Let LogicMonitor pull read-only agent-spend metrics, verify poll health, and preserve cold-start and null semantics.

Create, view and revoke meridian:ingest keys for LiteLLM, Kong, Portkey and capacity feeds.

API reference

Monthly event ceilings, soft warnings, hard stops, per-key rate limits, producer health and retention for OASA ingest.

Scheduled spend and usage pull from the OpenAI Usage API into Meridian spend records.

Scheduled spend and usage pull from Anthropic Admin Cost/Usage APIs.

Credits-level spend sync from OpenRouter. No per-model or token breakdown.

Spend from Azure Cost Management for Azure OpenAI deployments (tokens not included).

Spend from Google Cloud BigQuery billing export filtered to Vertex AI. Prerequisite: BigQuery billing export on the GCP project.

A daily Amazon Bedrock total from AWS Cost Explorer (tokens not included, no tags). For attribution, upload the AWS FOCUS-format Cost and Usage export instead: Meridian reads cost-allocation tags from application inference profiles, and when a cost center tag matches a department in Meridian, that department becomes the cost center on each charge whether or not an agent tag matches. An Agent tag that matches an agent name also sets the agent. Untagged usage lands in the attribution inbox with the reason shown. Prior-period corrections keep their original period and become adjustment lines once that month is locked. While the connector is active, Bedrock rows in a FOCUS upload are held for reconciliation.

AI Studio key validates the connection. Prerequisite: BigQuery billing export on the paid-tier GCP project for spend metering (free-tier has no spend to meter).

Upload PDF, CSV, or Excel invoices, or enter a bill manually. Formats processed into spend records.

Credit-level agent attribution for Copilot Studio — valued at your rates, reconciled to your Azure invoice.

Docs-only setup: add the open-source OASA exporter to your OpenTelemetry Collector and it sends one usage record per GenAI span to Meridian, from any agent instrumented with the OpenTelemetry GenAI conventions. No in-app Connect wizard.

Learn how to implement an edge proxy for AI APIs: route traffic, balance load, enforce policies, and cut latency. This Onaro™ guide covers architecture patterns, provider configuration, and safe rollout for high-volume OpenAI and Anthropic workloads.

2-4 hours · Intermediate

Add circuit breakers around LLM calls to stop cascading failures, shed load during outages, and avoid runaway spend when APIs degrade. Step-by-step patterns for retries, fallbacks, and observability in production AI systems.

2-3 hours · Intermediate

Implement semantic caching so similar prompts hit a cache instead of the model—often cutting API cost dramatically. Covers embeddings, similarity thresholds, invalidation, and when caching is safe for your use case.

4-6 hours · Advanced

Route tasks to the right model tier: cheap models for simple work, premium models where quality matters. Practical routing rules, evaluation tips, and examples to lower spend without surprising regressions.

2-4 hours · Beginner

Compress prompts and context to cut token usage 30–50% while preserving answer quality: summarization, structured extraction, trimming policies, and measurement so savings show up in your real traffic.

3-5 hours · Intermediate

Stream model responses to users for snappier UX without raising token cost. Covers SSE patterns, client handling, backpressure, and provider-specific streaming options for chat and agent interfaces.

1-2 hours · Beginner

Batch LLM and embedding jobs to unlock provider batch discounts and simpler rate limits. When to batch, how to chunk inputs, idempotency, and monitoring so throughput goes up and per-token cost goes down.

3-4 hours · Intermediate