See what your AI agents are actually doing
Every demo on this page runs against the live curate-me.ai production gateway. Real scans, real containers, real costs.
Instant
Results in under 1 second. No container boot, no signup.
Gateway Playground
Paste text and scan it through the live governance pipeline. PII detection, injection analysis, and security scanning.
Governance Walkthrough
Animated 6-step governance chain: rate limiting, cost estimation, PII scan, security scanner, model allowlist, and human approval.
Interactive deep-dives
HITL Auto-Replay
patent-pendingWhen a request exceeds cost/PII thresholds, the gateway queues it for human approval, then automatically replays the exact request, no caller resubmission needed. Zero competitors have this.
Request Received
Governance: HITL Triggered
Awaiting Approval
Human Approves
Auto-Replay
Response Delivered
Pre-Flight Cost Governance
patent-pendingEvery LLM call is estimated, budget-checked, and governed before tokens are consumed.
4-Level Budget Hierarchy
# Response headers added by gateway
X-CM-Request-Id: req_a7b3c9d2
X-CM-Cost-Estimate: 0.003
X-CM-Governance-Time-Ms: 3.1
X-CM-Trace-Id: 00-4bf92f3577b34da6a3ce929d0e0e4736
RateLimit-Remaining: 58
Watch it work
AI agents execute in secure sandboxes with desktop streaming and real-time cost tracking.
Explore
Self-improving systems and multi-agent orchestration.
Night Owl
A daily AI digest that improves with feedback. Rate editions and watch quality trend upward.
Chat
Multi-agent orchestration with 14 tools. Delegate to researchers, writers, analysts, and developers.
Full Dashboard
Open the command center with live work ledger, cost tracking, governance events, and runner management.
Everything you see is running on the curate-me.ai platform.
One URL change gives your agents the same governance chain, cost tracking, and audit trail. Free tier includes 1,000 gateway requests per day.