Deploying to your team

You deploy an analytics agent by putting a chat interface in front of your business users. Once the context is built and the tests pass, the agent is ready for people who do not write SQL. They ask a question in plain English and the agent runs the query, checks the result, and answers.

The data team builds the agent. The rest of the company uses it. That split is the point. The team owns the context and the users get answers without filing a ticket.

Where the agent lives

Business users do not want a CLI. They want the agent where they already work. With nao the chat UI ships as a Docker container and runs on the web, plus the messaging tools your team uses.

• Web. A chat page at your own URL.
• Slack. Ask in a channel or a DM and get the answer in thread.
• Teams. The same, inside Microsoft Teams.
• Telegram and WhatsApp. For teams that live in chat.

Meeting them where they are is what gets the agent used. An answer in Slack beats a dashboard nobody opens.

How non-technical users query

A user types a question the way they would ask a colleague. “How many trials converted last month?” The agent reads it, plans the query, runs it on the warehouse, checks the numbers, and replies in plain language. No SQL, no waiting on the data team.

Because the agent reads the context the team built, the answer uses your real metric definitions. Revenue means what finance says it means. The user does not have to know that, they just get a number that holds up.

What changes for the data team

The team stops being a queue. The questions that used to land as tickets now go to the agent. The team’s job shifts to owning the context that keeps the answers correct, plus the governance that controls what the agent can see. Deploying is not the finish line. It is the point where the agent starts earning trust, which means keeping it reliable becomes the ongoing work.