VCAL Project
VCAL Project/Documentation
VCAL Documentation
Documentation for the VCAL stack.
Deployment guides, configuration references, architecture notes and integration guidance for AI Cost Firewall and every VCAL module.
Start with AI Cost Firewall: it is the foundation, and the module integration guides are published in its documentation.
Fig. 01 — Start here
AI Cost Firewall documentation.
The OpenAI-compatible gateway for LLM traffic: routing, caching, limits, observability, production safeguards and orchestration of the VCAL modules.
- Quick start and Docker Compose setup
- Configuration and provider reference
- Exact and semantic caching
- Privacy, security, usage-policy, Audit and Compliance integration
- Observability, evidence and failure handling
AI Cost Firewall is open source. Source code, issues and releases are on GitHub.
Fig. 02 — VCAL modules
Module integration guides.
The commercial modules extend AI Cost Firewall. Their integration documentation is published in the AI Cost Firewall documentation set.
- Privacy GuardDetection, redaction, anonymization, restoration, configuration and failure behavior.Open guide →
- Security GuardRequest and response decisions, enforcement, block-response modes and observability.Open guide →
- Usage GuardAcceptable-use policy, tenant and policy selection, block behavior and metrics.Open docs →
- AuditEvidence delivery, trace reconstruction, retention, verification and backpressure.Open guide →
- ComplianceControl evaluation, findings, review scope and reporting boundaries.Open docs →
Fig. 03 — Specialized infrastructure
Semantic Cache and vcal-core.
Separate documentation for dedicated semantic-reuse infrastructure and for the open-source engine underneath it.
VCAL Semantic Cache
For deployments where semantic reuse must be shared, persistent or scaled independently of an AI Cost Firewall instance.
- Installation and licensing
- Docker and binary deployment
- Persistence, operations and observability
- API usage and production guidance
vcal-core
The open-source semantic matching library behind VCAL, for developers who need lower-level building blocks or direct integration.
- Core concepts
- Reusable building blocks
- Direct integration guidance
Fig. 04 — Can’t find it?
Ask us about your deployment.
For questions the docs don’t answer yet, about configuration, architecture or a specific environment, contact VCAL. Product-level answers are in the FAQ.