AI Security Gateway
An OpenAI-compatible proxy that adds a security and governance layer between your application and every LLM provider — without changing a line of application code.
PII Redaction
30+ built-in entity types — SSNs, emails, credit cards, API keys (OpenAI, Anthropic, Google, Groq, AWS), GitHub tokens, IBANs, crypto addresses, street addresses, and more. Detected and redacted in under 50ms.
Prompt Injection Blocking
Heuristic pattern engine covering jailbreaks, DAN variants, role overrides, SYSTEM OVERRIDE impersonation, ignore-previous attacks, and encoding exploits. Fail-closed model with sub-300ms gateway-level blocks.
Spend Enforcement
Per-project token budgets with hard 402 stops before overspend. Agent loop protection — no more overnight runaway costs. Different budget envelopes for dev vs. production.
Smart Cost Routing
Routes each request to the cheapest provider that meets your latency and capability requirements. 300+ models across OpenAI, Anthropic, Google, Groq, Mistral, and more. 40–60% savings in production.
Metadata Audit Logs
Every request logged with provider, latency, DLP latency, PII flag, cost, and token counts — without storing prompt content. Governance visibility without privacy trade-offs.
OpenAI-Compatible
Works with every OpenAI SDK, LangChain, LlamaIndex, and any framework that accepts a base_url. No middleware to write. Change one line, get the entire security layer.
Free tier — 1M credits · No credit card required · Apache 2.0 open source
Also From Datum Fuse
Our data intelligence platform turns complex datasets into clear decisions — automatically.
DatumFuse.ai
An AI-powered analytics platform that automates the entire data workflow — from cleaning and merging messy files to generating charts, narratives, and SEO keyword intelligence with one click.