YeshID

"Why does anyone your AI agent?" I am often asked. "This should've been an MCP" they tell me. Wrong, and I can prove it. See, it's all fun and games until you ask the agent to help you investigate and it refuses, or worse: ignores on purpose. We've already seen malwae randomly sprinkle the code with requests purposely designed to trigger most LLMs to skip (by embedding prompts to design ☢️/🧬 🔫). A good agentic outcome is determined by 4 things: 1. Model 2. Harness/tools 3. Context 4. Eval data I am building an agentic security system that will not refuse, will not get distracted, and will not rest.

We are about to repeat the SaaS sprawl mistake, except faster and with software that can act on its own. SaaS sprawl gave us dozens of app consoles, each with a partial answer about who has access to what. Nobody owned the whole picture. Access reviews became theater. Agent sprawl is the sequel. Every platform will give you a clean inventory of the agents it knows about. That data is genuinely useful. But the risk was never inside one console. It's in the relationship between the agent and the rest of your access model: the owner, the key, the app, the approval, the lifecycle. Agent inventory is becoming table stakes. Connecting those agents to who's accountable and whether they should still have access is the actual work. Wrote up the inventory vs. governance line here. https://lnkd.in/g7yUjRJ5

Somewhere in your company, someone left six months ago. The directory says they're gone. Their account is disabled. Offboarding "complete." But an agent they spun up is still running. It's still holding an API key. It's still reaching into systems. And the person who could tell you why it exists doesn't work here anymore. That gap, between what your directory believes and what's actually still running, is where identity risk lives now. Agents make it worse, because they chain tools, reuse credentials, and keep going long after the business reason is gone. Inventory tells you the agent is there. It won't close this gap. Governance has to. New post on the difference, and why it matters more every quarter. https://lnkd.in/g7yUjRJ5

We shipped something rough on purpose. YeshID Cost Management is live, and it's V1. We put it out early because we'd rather your feedback shape where it goes than guess in a room by ourselves. What it does today: all your SaaS and AI spend in one place, upcoming renewals with an audit trail, wasted seats you can right-size before renewal, and orphaned accounts you're still paying for that aren't tied to an active employee. Finance gets a dedicated Cost Management role, so you can give them spend visibility without opening up the rest of your identity data. Least privilege, even for the budget conversation. It's useful today. Tell us what's missing. That's the whole point of shipping now. https://lnkd.in/gc_K9PcA

We shipped agent inventory. Then I realized we'd answered the easy question. Knowing what non-human actors exist in your company is useful. It's also the part everyone is racing to do right now. "Look how many agents we found" makes a good screenshot. But it doesn't answer the questions security and IT actually have to defend: Who owns this agent? Who approved it? What can it reach? What credential is it using? What happens when the owner leaves? A list can't tell you whether access still makes sense. That's not inventory. That's governance, and it's the harder, less screenshot-friendly half of the job. Wrote down where the line is between the two. https://lnkd.in/g7yUjRJ5

NHI HI Why not both?

Being the only IT person at a growing company is one of the hardest jobs nobody talks about. That's where we come in. YeshID automates the Identity work that eats your day, so onboarding takes minutes instead of hours, offboarding actually finishes, and you can answer "who has access to what" without digging.

Your AI spend is the fastest-growing line item nobody's watching. That's not a coincidence. It's the same blind spot as the rest of non-human identity, now showing up on the invoice. YeshID Cost Management lets you connect OpenAI and Anthropic and pull real usage costs into the same view as your SaaS spend. Humans and non-humans, access and cost, in one place. You can't govern what you can't see. You also can't budget for it. Cost Management is live now. https://lnkd.in/gc_K9PcA

Your orphaned account is a money leak and an open door. Usually it's the same account. That's the idea behind YeshID Cost Management, which is live now. To govern access, we already know who can reach what across your SaaS, OAuth grants, service accounts, and AI agents. That same data shows exactly where you're paying for access nobody uses. Wasted seats. Shadow apps. Licenses tied to people who left in March. The waste and the exposure live in the same gap: between what you think you have and what's actually true. We're not becoming a vendor management company. We just couldn't keep looking at where the money was leaking and not show you. Open the Cost tab to get started. https://lnkd.in/gc_K9PcA

To everyone living with Oktholm Syndrome: You are not alone. There are many who suffer. We see you defending the renewal quote. We see you paying the SSO tax with a straight face. There is a cure. You don't have to live like this.