Console
AlchemiStudio’s Console is the developer control plane — the place where engineers build, test, and operate AI workloads. From the Console you can manage the full lifecycle of agents, workflows, prompt templates, datasets, guardrail pipelines, and production releases, and monitor everything through request traces, usage dashboards, and model health views.
What you can do
| Action | Description |
|---|---|
| Manage projects | Create and configure projects to namespace agents, workflows, keys, and deployments |
| Manage agents | Create gateway-bound agents with models, update agent configuration, and delete agents |
| Build workflows | Create and configure multi-step AI workflows |
| Manage prompt templates | Author, version, and publish prompt templates |
| Datasets & experiments | Build evaluation datasets and run experiments to measure prompt and model quality |
| Guardrails | Define guardrail policies (PII, jailbreak, toxicity, content filtering, secret detection, topical) and configure guardrail pipelines |
| Releases | Create and manage production releases; promote builds through environments |
| Virtual keys | Issue and manage virtual API keys scoped to a project |
| Integrations | Connect and configure third-party services available to agents and workflows |
| Monitor model health | View endpoint health status across all model providers |
| Inspect traces | Drill into individual request traces for debugging |
| View logs | Browse and filter the request log stream |
| Playground | Test agents and model targets interactively without writing code |
| Usage | Review token consumption, cost, and request volume across your projects |
Where to find it
Open AlchemiStudio and select Console from the product switcher in the left sidebar.
In this section
- Explore the Console Overview — tour of the Console dashboard and key metrics
- Create New Project in Console — setting up a project namespace
- Creating Gateway-Bound Agents with Models — registering agents against model endpoints
- Update an Agent — editing agent configuration
- Delete an Agent — permanently removing an agent
- Create and Configure a New Workflow — building multi-step AI workflows
- Creating Prompt Template — authoring and versioning prompts
- Managing Prompts — managing prompt lifecycle
- Create and Configure a Dataset — building evaluation datasets
- Experiment for a Dataset — running evaluation experiments
- Managing Evaluators — configuring evaluation metrics
- Create and Configure Guardrail Definitions and Pipelines — setting up guardrail policies
- Create a New Release — cutting a release from a build
- Managing Production Releases — promoting and rolling back releases
- Managing Virtual Keys in Projects — issuing and revoking API keys
- Connecting Third-Party Apps in Integrations — wiring in external services
- Monitoring Model Health Status — checking endpoint health
- Inspecting Request Traces — debugging individual requests
- Viewing Request Logs — browsing the log stream
- Playground Testing for Targets — interactive testing without code
- Viewing Console Usages — reviewing consumption and costs
Last updated on