Use Cases
Discover how STOA Platform solves real challenges for enterprises adopting AI and managing APIs at scale.
🤖 Use Case 1: AI Agent API Gateway
Challenge: Your AI agents (Claude, GPT, custom LLMs) need to call internal APIs, but you lack governance, audit trails, and access control.
Solution: STOA's MCP Gateway provides a unified entry point for all AI-to-API interactions.
Benefits
| Without STOA | With STOA |
|---|---|
| Direct API access, no governance | Centralized policy enforcement |
| No visibility on AI calls | Full audit trail per agent |
| Manual API key management | Auto-discovery via MCP |
| Scattered rate limiting | Unified quotas per tenant |
Quick Implementation
# 1. Register your API in STOA
stoa api register --name billing-api --upstream https://billing.internal
# 2. Claude.ai automatically discovers it
# In Claude: "What billing APIs are available?"
# Claude calls: stoa_catalog(action="search", query="billing")
# 3. Agent subscribes and calls
# stoa_subscription(action="create", api_id="billing-api")
# Direct API call through MCP Gateway with full audit
📊 Use Case 2: AI Cost Management & Token Metering
Challenge: Multiple teams use LLMs, but you have no visibility on token consumption, costs spiral, and there's no chargeback mechanism.
Solution: STOA's AI Gateway tracks every token, provides real-time dashboards, and enables per-team quotas.
Benefits
| Metric | Before STOA | After STOA |
|---|---|---|
| Cost visibility | None | Real-time per team |
| Token waste | Redundant LLM calls | Reduced with semantic cache* |
| Budget control | Monthly surprise bills | Automatic quota enforcement |
| Provider lock-in | Single provider | Multi-provider failover |
Semantic caching can significantly reduce redundant LLM calls. Actual savings vary based on query patterns and use case.
Dashboard Preview
┌─────────────────────────────────────────────────────────┐
│ Token Usage by Team (January 2026) │
├─────────────────────────────────────────────────────────┤
│ Engineering ████████████████████░░░░ 2.4M tokens │
│ Marketing ████████░░░░░░░░░░░░░░░░ 800K tokens │
│ Support ████████████░░░░░░░░░░░░ 1.2M tokens │
├─────────────────────────────────────────────────────────┤
│ Total Cost: €1,240 | Cache Savings: €520 (30%) │
└─────────────────────────────────────────────────────────┘
🏢 Use Case 3: Multi-Tenant API Platform
Challenge: You're an ESN/ISV exposing APIs to multiple clients. You need isolation, per-client quotas, billing, and compliance (NIS2/DORA).
Solution: STOA provides native multi-tenancy with complete isolation and European sovereignty.
Tenant Isolation Architecture
| Aspect | Implementation |
|---|---|
| Network | Kubernetes NetworkPolicy per namespace |
| Data | Schema-per-tenant in PostgreSQL |
| Auth | Separate Keycloak realm per tenant |
| Secrets | Isolated Vault paths |
| Quotas | Independent rate limits |
Compliance Features
- NIS2 — Security incident reporting, access controls
- DORA — ICT risk management, resilience testing
- GDPR — Data residency in EU, audit trails
- European Hosting — EU data residency option available*
Legal jurisdiction depends on multiple factors. Consult your legal counsel for compliance advice.
Getting Started
Ready to implement these use cases?
STOA Platform is currently in private beta. Request access to deploy these use cases on your infrastructure, or explore the Architecture Overview to understand how STOA works.
- ☁️ SaaS
- 🏠 Self-Hosted
1. Request beta access → christophe@hlfh.io
2. Receive your tenant credentials
3. Sign in at https://console.gostoa.dev
4. Create your first API in the Console
1. Request beta access → christophe@hlfh.io
2. Receive Helm chart access
3. helm repo add stoa https://charts.gostoa.dev
4. helm install stoa stoa/stoa-platform -n stoa-system
Need Help?
- 📚 Quick Start Guide
- 💬 Discord Community
- 📧 Contact Sales for enterprise use cases
Claude is a trademark of Anthropic, PBC. GPT-4 and Azure OpenAI are trademarks of OpenAI/Microsoft. All other product names are trademarks of their respective owners.