InferLane uses the following third-party services to process customer data. We give 14-day notice of changes to customers on enterprise contracts.
| Service | Purpose | Location |
|---|---|---|
| Vercel | Application hosting, edge functions, cron jobs | US (iad1) |
| Neon | PostgreSQL database (serverless) | US East |
| Upstash | Rate limiting (Redis) | US East |
| AWS | Underlying cloud infrastructure (incl. email delivery via SES) and storage | US / ap-northeast-1 |
| Resend | Transactional email (waitlist, digests) | US (via AWS SES ap-northeast-1) |
| Stripe | Payment processing, billing, identity verification, payouts | US |
| PostHog | Product analytics (consent-gated, no PII) | US |
| Anthropic | AI inference (proxied on behalf of users) | US |
| OpenAI | AI inference (proxied on behalf of users) | US |
| AI inference (proxied on behalf of users) | US |
AI inference providers are only contacted when a user explicitly sends a request through the InferLane proxy, and the content of that request transits the provider the user selects. We do not send data to inference providers proactively. Users choose which providers to connect via their dashboard or environment variables. In addition to the named providers above, InferLane can route to other LLM providers a user connects; the provider that processes a given request is the one the user selects for that request.