Platform
Curate-Me is the governance layer for AI agents.
If you already have apps, prompts, agents, or automations in production, Curate-Me adds the parts that usually get built late and painfully:
- cost controls
- security and policy enforcement
- managed execution environments
- visibility into what actually happened
Start Here
Plain-English explanation for non-technical readers, buyers, and new teammates.
How It WorksRequest flow, runtime map, gateway path, runners, and the main system boundaries.
Technical WalkthroughHow Curate-Me looks to engineering, product, security, support, finance, and execs.
By RoleDeeper implementation details for the triple API architecture and control plane.
ArchitectureThe Short Version
Most teams start AI projects with direct provider calls, scattered secrets, and little visibility into spend or policy violations.
Curate-Me inserts a governed layer between your app and model providers, then adds managed runners and operator tooling around it.
# Before
OPENAI_BASE_URL=https://api.openai.com/v1
# After
OPENAI_BASE_URL=https://api.curate-me.ai/v1/openai
X-CM-API-Key=cm_sk_xxxThat single swap gives you:
- gateway governance
- request logs and cost tracking
- approvals for risky actions
- managed runners for agent execution
- a dashboard for operators and teams
The Three Product Surfaces
| Surface | What it does | Who tends to use it |
|---|---|---|
| AI Gateway | Governs every model request before it reaches OpenAI, Anthropic, Google, or other providers | engineers, platform teams, security |
| Managed Runners | Runs agent workloads in controlled environments with lifecycle, files, templates, and fleet controls | engineering, ops, automation teams |
| Dashboard | Gives humans a way to inspect usage, policies, health, costs, approvals, and runner activity | engineering managers, support, finance, execs |
Good Next Reads
- New to the product: How It Works
- Technical audience: Technical Walkthrough
- Role-specific rollout planning: By Role
- Need implementation detail: Architecture