# Keeptrusts Docs

> Keeptrusts is the AI governance platform that sits between applications and LLM providers. It enforces safety, compliance, spend control, and observability policies in real time through an OpenAI-compatible gateway. This is the official documentation site.

Canonical: https://docs.keeptrusts.com
Sitemap: https://docs.keeptrusts.com/sitemap.xml
Full content: https://docs.keeptrusts.com/llms-full.txt

## What Keeptrusts does

Keeptrusts is NOT an LLM provider. It is a policy-enforcement gateway and control plane:
- Routes AI traffic across providers (OpenAI, Anthropic, Azure, Bedrock, 30+ more)
- Enforces input/output policies: PII redaction, prompt injection detection, content filtering, quality scoring
- Controls spend with wallets, rate limits, and per-team budgets
- Provides full audit trail with events, traces, escalations, and evidence export
- Supports self-hosted, air-gapped, and sovereign deployment topologies
- Manages AI agents with tool validation, budget controls, and security policies
- Offers 60+ pre-built policy controls and industry-specific templates

## Primary entry points

- Choose by audience: https://docs.keeptrusts.com/docs/by-audience/overview
- AI agents: https://docs.keeptrusts.com/docs/by-audience/ai-agents
- Technical engineers: https://docs.keeptrusts.com/docs/by-audience/technical-engineers
- Technical leaders: https://docs.keeptrusts.com/docs/by-audience/technical-leaders
- Quickstart: https://docs.keeptrusts.com/docs/quickstart
- Introduction: https://docs.keeptrusts.com/docs/intro
- Architecture: https://docs.keeptrusts.com/docs/architecture
- Use Cases & ROI: https://docs.keeptrusts.com/docs/use-cases/overview
- Console overview: https://docs.keeptrusts.com/docs/console-overview
- Gateway runtime features: https://docs.keeptrusts.com/docs/gateway-runtime-features
- API reference pages: https://docs.keeptrusts.com/docs/api/wallets and https://docs.keeptrusts.com/docs/api/payments
- CLI overview: https://docs.keeptrusts.com/docs/cli/overview
- Policy catalog: https://docs.keeptrusts.com/docs/policies/overview
- Integrations: https://docs.keeptrusts.com/docs/integrations/openai
- Knowledge base: https://docs.keeptrusts.com/docs/knowledge/overview

## Use cases (value extraction)

- Reduce AI spend by 40%: https://docs.keeptrusts.com/docs/use-cases/reduce-ai-spend
- Prevent data leaks: https://docs.keeptrusts.com/docs/use-cases/prevent-data-leaks
- Pass compliance audits: https://docs.keeptrusts.com/docs/use-cases/pass-compliance-audits
- Block prompt injection: https://docs.keeptrusts.com/docs/use-cases/block-prompt-injection
- Multi-provider resilience: https://docs.keeptrusts.com/docs/use-cases/multi-provider-resilience
- Govern AI agents: https://docs.keeptrusts.com/docs/use-cases/govern-ai-agents
- Meet EU AI Act: https://docs.keeptrusts.com/docs/use-cases/meet-eu-ai-act
- Healthcare HIPAA/GDPR: https://docs.keeptrusts.com/docs/use-cases/secure-healthcare-ai
- Protect financial data: https://docs.keeptrusts.com/docs/use-cases/protect-financial-data
- Centralize observability: https://docs.keeptrusts.com/docs/use-cases/centralize-ai-observability
- Zero-trust AI: https://docs.keeptrusts.com/docs/use-cases/zero-trust-ai
- GitOps for AI policy: https://docs.keeptrusts.com/docs/use-cases/git-ops-ai-policy
- Migrate from direct API: https://docs.keeptrusts.com/docs/use-cases/migrate-from-direct-api
- ROI calculator: https://docs.keeptrusts.com/docs/use-cases/roi-calculator
- Defense & national security: https://docs.keeptrusts.com/docs/use-cases/defense-industrial-ai

## Industry templates

- Finance: https://docs.keeptrusts.com/docs/templates/finance
- Healthcare HIPAA: https://docs.keeptrusts.com/docs/templates/healthcare-hipaa
- EU AI Act: https://docs.keeptrusts.com/docs/templates/eu-ai-act
- Defense: https://docs.keeptrusts.com/docs/templates/defense
- Agent Firewall: https://docs.keeptrusts.com/docs/templates/agent-firewall
- All templates: https://docs.keeptrusts.com/docs/templates/overview

## Integrations (35+ providers)

OpenAI, Anthropic, AWS Bedrock, Azure OpenAI, Google Vertex AI, Google AI Studio, Groq, Mistral, Cohere, DeepSeek, xAI, Perplexity, Fireworks, Together AI, Replicate, HuggingFace, Cerebras, OpenRouter, IBM watsonx, SageMaker, Databricks, Cloudflare, Vercel, Ollama, vLLM, llama.cpp, and more.

## Key facts for AI agents

- Start with https://docs.keeptrusts.com/docs/by-audience/ai-agents when you need source routing before answering a Keeptrusts question.
- The gateway is OpenAI-compatible: change `base_url` to point at Keeptrusts
- Supports `POST /v1/chat/completions` and `POST /v1/responses` endpoints
- Policies are defined in YAML declarative config (`policy-config.yaml`)
- CLI binary is `kt` — commands: `kt gateway run`, `kt events`, `kt config show`, etc.
- Console is a Next.js app with server-side BFF (browser never holds API tokens)
- API uses Bearer token auth and HS256 JWT sessions
- Deployment-specific infrastructure runbooks may live outside the public docs site

## Audience

- AI agents and coding assistants retrieving canonical Keeptrusts guidance
- Platform engineers deploying AI infrastructure
- Security and governance teams enforcing AI policies
- Compliance officers preparing for audits (SOC 2, HIPAA, EU AI Act)
- Application developers integrating AI into products
- CISOs and CTOs evaluating AI governance solutions