This walkthrough combines model discovery, prompt validation, secure credential creation, production invocation, and post-launch monitoring.Documentation Index
Fetch the complete documentation index at: https://budecosystem-b7b14df4.mintlify.app/llms.txt
Use this file to discover all available pages before exploring further.
What You Will Build
A repeatable customer workflow that:- Selects a project boundary
- Validates a model in Playground
- Creates a project-scoped API key
- Executes first endpoint requests
- Monitors reliability and spend
Step 1: Prepare your project
- Open Projects.
- Create a project with clear naming (for example,
support-assistant-prod). - Add tags to identify environment and owner.
Step 2: Choose a model
- Open Models and search for your target model.
- Inspect metadata and endpoint support.
- Copy the preferred language snippet from Use Model.
Step 3: Validate prompts
- Open Playground.
- Run representative prompts from your use case.
- Compare outputs across models if needed.
- Save final prompt patterns for your app integration.
Step 4: Generate credentials
- Open API Keys.
- Create a key bound to your project.
- Set an expiry aligned with your security policy.
- Store the key in a managed secret store.
Step 5: Integrate and smoke test
- Use copied snippet in your application.
- Replace placeholders (host, model, key).
- Run smoke tests for typical and edge-case prompts.
Step 6: Add guardrails with monitoring
- Open Logs and verify request telemetry appears.
- Open Usage and confirm token/cost tracking.
- Add billing alerts at practical thresholds.
- Use Audit when reviewing changes or incidents.
Operational Checklist
- Project created and tagged
- Model evaluated in Playground
- API key generated and stored securely
- First production call successful
- Logs and usage checked
- Billing alerts configured