IQ Routing Client Cost Optimization Setup (Onboarding)

IQ Routing Client Cost Optimization Setup (Onboarding)

1. 01

   Create an IQ Routing account and select the Team plan for per-team budgets.

   Navigate to the IQ Routing dashboard and sign up. Choose the Team plan at $70/month to access per-team budgets, audit logs, and custom band maps, which are essential for managing multiple clients.

2. 02

   Configure the unified API endpoint with your client's provider keys.

   In the Settings screen, add your client's OpenAI, Anthropic, and Google API keys. IQ Routing provides a single URL for all three providers, so you can replace the base URL in the client's SDK without code changes.

3. 03

   Set up per-team budgets for the client.

   Use the Teams section to create a team for the client and assign a monthly budget. This enables cost tracking and alerting per client, preventing overspend and providing clear billing data.

4. 04

   Define custom band maps for routing policies.

   In the Routing Rules area, configure band maps (auto, cheap, frontier) to match the client's quality requirements. For example, set cheap models for simple queries and frontier models for complex reasoning tasks.

5. 05

   Enable semantic caching to reduce redundant API calls.

   Turn on the semantic cache in the Cache Settings. This avoids re-billing for repeated requests, which is common in chatbots and RAG systems, potentially cutting costs further.

6. 06

   Test the integration with a pilot workflow.

   Change the base URL in the client's SDK to the IQ Routing endpoint and run a few test requests. Monitor the dashboard to verify that routing works and that cost savings appear as expected.

7. 07

   Document the handoff for the client's team.

   Prepare a handoff document explaining the new endpoint, budget alerts, and how to read the cost dashboard. Include instructions for adjusting band maps if quality issues arise.