NyLi Doesn't Just Wait for You to Ask

Most people think NyLi only talks when you talk to it first

You open Atlas. You type a question. NyLi answers. That's the mental model most users carry: NyLi is something you ask, not something that speaks up.

But NyLi also queues up messages on its own. A daily briefing when you start your day. A nudge when a follow-up is coming due. An alert when a deal has gone stale and needs attention. A heads-up when a longer-running job—an import, a specialist agent's task—finishes in the background.

This isn't new. It's been part of NyLi for a while. But it's quiet enough that many users never notice it's happening. They see NyLi as reactive intelligence, not proactive watch.

The user should not have to work the software

A stale deal shouldn't require you to remember to go looking for it. An overdue follow-up shouldn't live in your mental stack until you manually pull a report. That's making the user do the software's job.

Atlas is built around the belief that intelligence should serve the work, not wait for permission to be useful. NyLi proactive messaging is the expression of that belief at the smallest scale: the system knows something worth mentioning, so it mentions it.

The harness isn't just connecting models to tools. It's connecting business signals to human attention at the right time.

Proactive messaging is unglamorous and deeply useful

Daily briefings aren't moonshots. Follow-up nudges aren't frontier AI. Job completion notices are reminders, not breakthroughs. This is operational intelligence—useful, quiet, easy to overlook.

But operational intelligence is where most work actually gets lost. You meant to circle back. You forgot the import was still running. You didn't realize a deal had been sitting untouched for two weeks until someone asked about it in a meeting.

NyLi proactive messages don't replace your judgment. They surface the context you need to apply that judgment without having to go hunting for it. That's the difference between intelligence and a harness around intelligence.

• Daily briefings orient your attention when you start working
• Follow-up nudges prevent dropped threads
• Stale deal alerts catch patterns you'd otherwise miss until too late
• Job completion notices close the loop on background work without requiring you to check back

If you've only been asking NyLi questions, you're missing half the system

Open Atlas tomorrow morning. Before you ask NyLi anything, check if it's already queued something up for you. If you've been working in Atlas for a while and you've never seen a proactive message, you might not have context tied in yet—or you might just be moving fast enough that nothing's had time to go stale.

But if you do see those messages, pay attention to when they show up and what they're surfacing. That's NyLi watching your business context and deciding what's worth your time. It's doing exactly what the rest of Atlas does at a larger scale: routing the right signal to the right person at the right moment.

Proactive messaging isn't a feature you turn on. It's a standing capability inside the harness, running quietly alongside everything else. You don't have to do anything to make it work. You just have to notice it's already working.