Admin Console Overview#

The Admin Console is the administrative workspace for configuring, governing, and monitoring OptScale AI.

Use it to configure providers and access, apply policies and guardrails, connect MCP servers, and monitor usage, cost, and request activity.

For how the Admin Console and Chat work together, see Architecture Overview.

For a new organization, start with First Steps. Otherwise, use the workflow below or go directly to the Admin Console menu.

Recommended setup flow#

Configure providers#

Start by connecting at least one AI provider. Open Config → Providers to:

• Add provider connections.
• Enable available models.
• Configure provider limits and settings.

Add Routing Rules when traffic requires fallback, weighted distribution, or other routing behavior.

Configure user access#

Open Config → Credentials & Roles to:

• Invite users.
• Configure allowed providers.
• Set spend or quota limits.
• Configure context compression.

For shared access and configuration settings, open Config → Teams. For application and agent identities, use Config → Agents.

Add governance controls#

Open Config → Policies & Guardrails when requests need safety, compliance, or other governance controls.

• Guardrails define reusable checks and actions.
• Policies define when those guardrails apply.

Add MCP integrations#

Add Config → MCP Servers when Chat or API workflows need approved external tools or data sources.

MCP configuration includes server connection settings, authentication, tool availability, and access for users or workflows.

Monitor and optimize#

After traffic starts flowing through OptScale AI, use the monitoring and analytics areas to review platform activity.

• Home — Review high-level cost, volume, reliability, and optimization trends.
• Analytics → Optimizations — Measure savings from optimization features.
• Analytics → Usage and Cost — Analyze token usage, spend, and activity by model, user, team, or agent.
• Analytics → Traces — Inspect individual AI requests, including routing, latency, token usage, cost, and payload details.
• ROI & Impact — Define and track business impact metrics and reports.

Admin Console menu#

The Admin Console sidebar groups features by purpose.

Home — View high-level operational, usage, cost, and optimization information.

Analytics

• Optimizations — Measure savings from optimization features.
• Adoption — Review organization-level AI adoption.
• Usage and Cost — Analyze token usage, spend, and scoped activity.
• Budgets — Define and monitor AI spending budgets.
• Cost Centers — Organize and review AI spend by cost center.
• Traces — Inspect and troubleshoot individual AI requests.

ROI & Impact

• Overview — Open dashboards and review metric results.
• Reports — Group metrics into reports and review their latest values.
• Metrics — Define what to measure and review current values.
• Schedules — Recompute active metrics automatically and optionally send a summary after each run.
• Signals — Review stored signals used as source data for analysis.

Config

• Credentials & Roles — Manage users, access, limits, allowed providers, and optimization settings.
• Teams — Manage team-level access and shared configuration.
• Agents — Manage application and agent identities and their access settings.
• Providers — Configure AI providers, models, and provider settings.
• Policies & Guardrails — Configure governance rules and reusable guardrails.
• MCP Servers — Connect approved external tools and data sources.

For routing configuration, see Routing Rules and Routing Best Practices.

For detailed governance configuration, see Guardrails and Policies.

How to Use

• Chat — Open the Chat access page and navigate to Chat.
• API & Agents — Find connection information for applications and agents.

System

• Settings — Manage system and organization configuration options.

See also#

• First Steps — Initial organization and provider setup.
• Architecture Overview — Platform structure and request flow.
• Interface Overview — Chat workspace for end users.