AI product
Stigg is a usage runtime for AI products that meters usage, manages entitlements, and enforces credits, limits, and budgets before an AI request runs. It provides a real-time enforcement layer for customers, users, and agents, with entitlement checks reported at p99 under 10 ms, usage ingestion at more than one million events per second, and support for more than one million root entities. Its credit infrastructure includes wallets, ledgers, burn-downs, expiry rules, and priority-based consumption, while governance features support spend caps and hierarchical controls. Stigg is offered as a hosted service or for deployment in a customer's own VPC, and is intended to handle usage enforcement and settlement without requiring the customer to build that billing infrastructure.
1 use taken from transcripts — each links to the moment in the video.
The speaker's company — Dor Sasson is its co-founder and CEO. His talk makes the case for the kind of financial infrastructure Stigg operates in: synchronous entitlement checks before inference, asynchronous settlement, hold-and-settle, credit pools, budgets and spend caps. He invites attendees to the Stigg booth where the team is demoing the product.
1 in the library.