Connect the systems where technology spend begins.
OggyCloud is being built around the tools modern teams already use: AWS, OpenAI, developer platforms, data infrastructure, SaaS subscriptions, and ownership metadata.
Start narrow. Build the data model correctly.
The first integration scope should prove the full workflow with AWS and OpenAI before expanding into every provider.
Cloud infrastructure
Start with AWS cost and usage data, then expand into Azure, GCP, Kubernetes, and managed platform signals.
AI providers
Capture OpenAI usage first, then support Anthropic, Gemini, model gateways, and agent workflow telemetry.
SaaS and platforms
Track spend from Vercel, GitHub, MongoDB Atlas, observability, data tools, and other operating platforms.
Ownership sources
Use tags, teams, repositories, projects, cost centers, environments, and metadata to route spend correctly.
Manual imports
Support CSV and invoice imports for tools that do not expose billing APIs or are not connected yet.
Custom events
Accept usage events from internal apps, AI gateways, and workflows when native billing data is not enough.
A practical integration path for early pilots.
These labels describe intended product scope, not a claim that every connector is already production-ready.
Each connector should feed one unified spend model.
The goal is not to copy every provider console. The goal is to normalize enough data to make ownership, alerts, and recommendations useful.
Read-only access
Connect billing and usage APIs with the narrowest practical access for each provider.
Normalize records
Convert provider-specific invoices, services, keys, and resources into one spend model.
Attach ownership
Map spend to projects, teams, repositories, workflows, cost centers, and environments.
Route action
Send alerts and recommendations to the owner with context, impact, and evidence.
Not every source needs a full connector on day one.
For the first pilots, OggyCloud can combine direct API connections with structured imports and manual mapping. That keeps the product useful while the connector library matures.
Have a stack we should support first?
Early pilot conversations will decide which connectors matter most after AWS and OpenAI.