---
title: "We ship website changes faster than we can watch - Atlan Frontier Labs"
url: "https://atlan.com/frontier/essays/one-agent-never-stops-working/"
description: "How one AI agent became the first thing on our team that never stops checking atlan.com. Optimus watches every page, every day - nothing waits for a human to notice, and nothing ships without a human deciding."
---

> Atlan is hosting Context Conference, bringing together the leaders and builders at the frontier of giving AI the context it needs to understand their business. It runs online on October 28, 2026, from 11:00 AM to 2:00 PM ET. Atlan co-founder Prukalpa Sankar opens and closes the day. Leaders from AstraZeneca, BNY and Verizon share why they invest in context and what they get from it. Registrants get early access to The AI Context Gap, a new study from MIT Technology Review Insights. Register: https://atlan.com/context-conference/

An essay from Atlan Frontier Labs (essay 002) by Dubem & Kushagra, Website operations, Atlan, dated 2026.07.28, 1,527 words. Title on the page: "We ship website changes faster than we can watch. So one agent never stops working." How one AI agent, Optimus, became the first thing on the team that never stops checking atlan.com: it watches every page, every day; nothing waits for a human to notice, and nothing ships without a human deciding. Part of [Becoming Frontier](https://atlan.com/frontier/).

## One evening last spring

One of the authors opened atlan.com on his phone and watched a blank rectangle sit where the homepage should be: eight seconds of nothing, then close to twenty more seconds where the page ignored every tap while the browser worked through styling the homepage never needed. Error console: green. Alerts channel: quiet. Performance test: 31 out of 100. The site was failing in front of anyone who opened it, every instrument said everything was fine, and the only way the team would have found out is a customer telling them.

Nothing errored because nothing broke the way software usually breaks. The site is built so a content update never touches code: pages are composed from a library of pre-built sections, and the build system decides automatically which styling each page needs (that automation lets a small team run thousands of pages). A styling bundle for a few new sections was labelled in a way that slipped past the rule deciding what is shared, so it began loading everywhere, the homepage among them. No single system was wrong; the failure lived in the space between three correct ones, so no alarm could point at it. "Once we knew, the fix took minutes. But knowing was the hard part."

## When we decided to reimagine everything

That gap is why Optimus exists, and where the team parts ways with the usual AI pitch:

- The usual pitch: efficiency, do the same work faster.
- What we build for: AI that owns outcomes a human team cannot hold at all, however fast it works. Keeping every page on atlan.com fast, for real visitors, every day, through problems that trip no alarms, is one of those outcomes, and the goal is to outperform.

The homepage was not a fluke; it is the daily condition of a site that ships faster than any person can watch it. As of last spring almost none of the pages passed Google's Core Web Vitals on mobile: over 1,100 sat in "needs improvement", effectively zero were green, and a single March regression dropped nearly 950 pages into the red overnight. The bottleneck was never the fix (often one line) but noticing it in time on the pages that matter: hundreds of URLs, vitals that move daily on real-user data, every deploy a candidate regression, every diagnosis hours in DevTools, commit logs, and analytics.

## What Optimus does

Optimus is an AI agent that lives in Slack. It watches every page on atlan.com worth watching, every day, without being asked. Its daily loop:

1. **Starts smarter than yesterday.** It keeps a running notebook (a page to watch after a recent change, a "broken link" that is really an external site blocking us) and reads yesterday's notes first, so each day starts where the last ended.
2. **Checks what a visitor would feel.** It walks every tracked page and measures what a visitor feels rather than what the console reports: the blank seconds, the layout that shifts underfoot, the tap that goes nowhere, across the real devices and networks visitors use. The team keeps feeding it more signals (search, traffic, conversion) so a slow page registers as costing something, early enough to act on.
3. **Traces the cause.** It traces each problem back to the recent code change most likely responsible (say, a tweak to a shared component that touched dozens of pages). That is a lead for the fix to start from.
4. **Files the fix and the guardrail.** For every issue it writes how to fix it, with enough detail that Mothership can implement it without coming back for context, and a guardrail: a check that would have caught it before it shipped, so the same class of mistake cannot slip through twice.
5. **Keeps its watchlist honest.** From the day's Slack chatter it adds pages that now matter (a new launch, a page suddenly in focus) and retires ones that don't, then leaves notes for the next run. That notebook is its memory across days.

Two to three days after a fix ships, Optimus re-checks the real-world data to confirm the improvement landed.

## The human gate: nothing ships without a decision

Optimus's autonomous surface is read-only; it changes nothing in production. On its own authority it:

- queries analytics and runs diagnostics
- reads source code and correlates commits
- files tickets and posts findings to Slack

A figure shows a real morning report in Slack (July 2026): the sweep, a live 404 traced to its commit, two tickets filed to Mothership.

Code changes run downstream through Mothership, Atlan's AI development platform that reviews and gates every engineer's code (more on it in a future piece). Mothership implements the fix and opens a pull request; a human reviews and merges it, the same gate every engineer's code passes through. Every fix ships with a preview build linked in Slack before merge, and the loop closes only when post-merge real-user data confirms the improvement days later. A second figure shows one real fix end to end: Optimus's report flagging render-blocking CSS leaking onto the homepage, the ticket delegated to the Mothership agent, an engineer's AI-assisted review, the pull request, and the merge.

How the team describes the split: "All of the claws are basically waiters in a restaurant. They go and talk to the chef in the back kitchen. The chef is the Mothership, taking care of everything." Optimus takes the order, carries it to the kitchen with full diagnostic detail, and brings the result back. Optimus does the thinking; Mothership does the building. The team's version of "autonomous" is narrow, and they argue stronger: nothing waits for a human to notice, and nothing ships without a human deciding.

## The impact

- **94% off the page weight, no visible difference.** A content update shipped three uncompressed images: 4.7 MB and a 25-second mobile load. The daily sweep caught it the same day and cut the weight by 94%.
- **A thousand-plus pages faster, from a two-line fix.** Optimus traced a sitewide drag to two lines in a shared template that roughly 1,340 pages inherit, named the exact file and lines; deleting them fixed all 1,340 at once.
- **A launch protected before its audience arrived.** A brand-new page went out heavy, was caught the next morning, and was fast again before its visitors showed up.
- **Six seconds of load time recovered with one line.** An article page lost six seconds to a single eager video embed.

Every one of these was invisible to the error console, and each would have cost weeks of drift before anyone connected it to its cause.

Total blocking time (the JavaScript-execution proxy) sat above one second for months, then dropped under the 200ms bar in the second week of May, after the agent went live (Lighthouse lab measurement across 13 tracked pages, not field data; from Optimus's own readout, internally called WebsiteClaw, May 2026). The fixes: deferred non-critical loaders, tracking moved out of the critical path, less hydration on content pages; about 1.1 seconds of blocking time removed. The operational queue (check this page, diagnose that regression, chase that fix) moved off people's plates.

Results by 20 July 2026: 1,516 mobile pages passed Core Web Vitals, 18 needed improvement, and zero were "poor", on both mobile and desktop. Desktop performance score went from 57 to 88; total blocking time from over a second to under 200 milliseconds. Fixes an agent opened, a human merged. (Thresholds are Google's, on real-user field data at the 75th percentile: "poor" means LCP of 4.0s or more, CLS of 0.25 or more, or INP of 500ms or more. Source: Search Console's Core Web Vitals report, backed by the Chrome UX Report. Before-snapshot early April, after-snapshot July 20.)

Kushagra: "A lot of work has been delegated, which basically means we focus on the bigger, better things rather than working on operational stuff. You basically teach AI how to get the job done once or maybe twice. Then you focus on how external stakeholders are experiencing pain points, fix the process, make it more seamless so everyone moves faster, rather than you just executing." The mental model shifted from "I do the work" to "I teach the system how to do the work, then improve the system."

## What we're still figuring out

- **Latency.** Optimus spins up a virtual instance, pulls in context, and understands the query before it starts working. The team is working to make it faster.
- **Long-term memory.** The day-notes notebook works (a hunch on Monday becomes a check on Tuesday), but Optimus cannot yet learn from a mistake once and never repeat it weeks later, unprompted. An agent that genuinely gets better each week is the next hard problem.
- **Catching problems before they ship.** Today it catches most problems after they happen, though before any human notices. The next phase: finding UI and UX issues before anyone reports them and delivering the diagnosis with a proposed fix attached, with humans still owning every merge.

## Final thoughts

The team did not set out to do the old work faster; they set out to own an outcome no human team could hold. It is early days, but Optimus is starting to get there. The site now holds a standard it never held before: fast on every page, every day, without depending on someone remembering to look, and with a human on every change that ships. "We got there by handing an outcome to a system built to own it. That is the version of AI-native worth building toward."

## Authors

Dubem & Kushagra, Website operations, Atlan. Dubem and Kushagra run atlan.com and built Optimus, the agent that never stops checking it. This piece carries their fixes, their misses, and their notebook.

## Read next

- [We rebuilt our content engine four times](https://atlan.com/frontier/essays/a-self-learning-content-engine/) (2026.08.25, 12 min read)
- [What it took to build an agent our CSMs actually trust](https://atlan.com/frontier/essays/an-agent-our-csms-actually-trust/) (2026.09.18, 11 min read)
- [Becoming a frontier company, in the open](https://atlan.com/frontier/essays/becoming-a-frontier-company-in-the-open/) (2026.07.23, 7 min read)