Private AI 26 May 2026 12 min read
Put one assistant where the questions already arrive, let it search only what each person may read, and measure how fast people reach a verified answer.
Read the guide ↗ One assistant in Slack in front of every company source A question asked in a Slack thread reaches one assistant. The assistant passes through a gateway that applies budgets, limits and guardrails, then reaches three separate sources, each with its own lock: policies, the help desk, and code and docs. The answer returns to the thread with citations. Conceptual illustration. Slack thread Where's the current expense policy? Policy v3, updated in March. 2 citations One assistant Gateway budget, limits, guardrails Policies Help desk Code and docs One front door, many sources Slack supplies the identity, the gateway applies budgets and guardrails once for every agent, and each source keeps its own access list.
Private AI 5 May 2026 13 min read
Give each team governed access to a few approved models through one gateway, and measure adoption by the work that gets finished.
Read the guide ↗ Four teams reaching four approved model routes through one gateway Support, sales, finance and engineering each hold their own key and connect to a central gateway. The gateway checks the key, the budget and the allowed routes, then passes the request to one of four routes: routine, hard cases, documents and private. This is a conceptual illustration. Each team, one door, four routes Support Sales Finance Engineering Gateway Own key per team Budget and routes One usage log Routine open weight Hard cases frontier Documents retrieval Private self-hosted Provider credentials stay inside the gateway. Teams only ever hold their own key. Governed access is one door with several rooms A team's key says which routes it may use and how much it may spend. Adding a team is issuing a key. Removing one is disabling it, and every application that used it stops at once.
