Olly says Hi: Scheduled tasks now report to Slack and email

An agent that only speaks when spoken to is a tool you have to remember to use. Olly has run on a schedule for a while now, working a saved prompt hourly, daily, weekly, or monthly and writing its findings into a chat with its own run history. Those scheduled tasks are now wired into the Coralogix Notification Center, so Olly delivers that output itself, allowing Olly to reach out to Slack or email, out of the box.

Giving the agent a delivery path of its own

Stanford’s 2026 AI Index puts enterprise AI adoption at 88% of organizations using AI in at least one business function, while agent deployment sits in the single digits across nearly every business function. The capability side of that gap is closing quickly: the same report has agents succeeding on 66.3% of real computer tasks, up from 12% in early 2024. What separates a capable agent from a deployed one is usually the wiring around it.

The Coralogix Notification Center already solves this problem for everything else in the platform. Routers do label-based matching on incoming notification requests and decide where they go. Now, scheduled Olly tasks work in the exact same way. Olly says hi!

Reports you can try today

Alert-shaped notifications already have a path. Olly’s automatic case analysis posts its root-cause finding onto the case and routes it to that case’s existing destinations without any per-channel setup. Scheduled delivery covers the other kind of reporting, the recurring cross-cutting read on how the estate is behaving, which today usually exists as a dashboard a team walks through in a weekly meeting.

A weekly incident review from your cases

Cases represent the operational problem rather than the symptom, and automatic case analysis has Olly investigate matching cases as they open, posting an investigation summary, a probable root cause with a confidence level, and two to four ordered remediation steps onto the case’s Triage tab. Filtering runs on case priority and ownership labels, so you decide whether Olly picks up every P1 or every case labeled with the payments service.

That leaves a week of investigated cases sitting in the platform. A scheduled task reading across them produces the review a lot of engineering organizations write by hand every Monday: which incidents opened, what caused them, how quickly each was acknowledged, which are resolved, and which are still active. Route that to a leadership channel and the weekly resilience conversation stops depending on whoever had time to compile it. Ask for it in the terms you want to read: “summarize every case opened in the last seven days, grouped by service, with the root cause Olly found, the time to acknowledgment, and a separate list of anything still open.”

Weekly delivery reports

Most operational surprises trace back to a change. Coralogix tracks those through event markers, which put every dated change on the same timeline as your service metrics. Deployments arrive as cdEvent payloads from GitHub, ArgoCD, Jenkins, and other pipeline systems, with build, test, and deploy metadata attached. Anything else, from a feature flag rollout to a marketing campaign, arrives as a customEvent. Both land through the ops-events ingest API at POST /v1/ops-events, and both are stored as opsEvents entities in the default/ops.events dataset, so DataPrime queries them like any other telemetry.

Once that data is queryable, a weekly deployment digest is a prompt rather than a project. Olly reads the week’s ops events, groups them by service and environment, and tells you what went out, from which pipeline, and whether anything landed close to an incident. DORA’s 2025 report makes the case that AI acts as an amplifier, magnifying whatever an organization’s delivery system already does well or badly. The teams shipping faster with AI assistance need a change record that keeps pace, and a Friday afternoon summary of every deployment across the estate is a cheaper way to keep it than a hand-maintained spreadsheet.

Operational health trends for your infrastructure and applications

Four hundred alerts fired this week means nothing until you know that last week it was ninety, and that three quarters of the increase came from one service.

Comparison is what makes a health report worth reading, and it is the kind of multi-step read the autonomous agent handles well. A weekly task counts alerts fired against the previous week, does the same for error log volume, names the services that appear in both windows, and flags anything trending in the wrong direction. Services that keep showing up week after week can rightfully have time invested in them. Not only does this surface key insights, it prevents tribal knowledge from functioning as a bottleneck in your SRE and Platform teams.

Cloud spend and FinOps analytics

Cost is telemetry, and the AWS Billing integration brings billing data into Coralogix alongside everything else, with dashboards and alerts for current costs, trends, and savings opportunities. Because it lands as ordinary queryable data, Olly can read it the same way it reads a latency metric, and produce FinOps reports of your entire AWS estate, with minimal token usage and a tiny data footprint, thanks to DataPrime serverside aggregation and the efficiency of the AWS Billing integration.

Setting one up

Scheduled tasks live at Settings, Olly AI, Scheduled tasks. A task takes a name, the prompt phrased the way you would phrase it in conversation, a model, a recurrence of one-off, hourly, daily, weekly, or monthly, and a start time that must be in the future and no more than a week out. The active toggle decides whether it runs immediately or waits. Connectors attach during creation or afterward, and you can also just tell Olly where to send the results.

Write the prompt to produce something clear, with special focus on any key insights you wish for Olly to emphasise. Say who it is for and what shape you want the answer in, and Olly will write to that rather than to the person who set the task up. The scheduled tasks documentation covers the form, the run history, and the connector setup.

This represents a further expansion of Olly’s abilities, and a deeper integration into the wider Coralogix platform, as we continue on our mission to expand Coralogix’s position as the only platform that was built for AI, not repurposed for it.