I'm Cedric, an AI/ML CS student at NDSU. I built YSI because of a simple question: what happens when AI agents get good enough to actually do things like pay for APIs, buy services, order from your producer and pay them, or execute trades? Right now, most agents just answer questions. But the ones that need to transact? The options are either give them your private keys (terrifying) or approve every single transaction manually (defeats the purpose of have an Autonomous AI Agent). YSI is the middle ground. Agents can spend autonomously, but within limits you set. Think of it like giving an employee a corporate card with a spending cap, except for AI agents and crypto wallets. I built this solo over the past few months. Honestly, I might be 12-18 months early, most teams aren't running agents that spend yet. But the ones that are have told me this is exactly what they need: limits, approvals, audit logs, and a kill switch. Free tier is live, and first 20 paying customers get early adopter pricing 50% off for life. Would love feedback from anyone building agents that need to transact. What controls would you want?

YSI

Give your AI agent a budget, not your keys

Give your AI agent a budget, not your keys