Immersive intelligence for smart and sustainable enterprises - the quiet software platform turning industrial operations into a live conversation with energy, water and carbon.
It is a Tuesday afternoon inside a hyperscale data center somewhere east of San Francisco. Outside, the AI boom is being narrated in press releases. Inside, the chillers are doing the actual narrating - and a piece of software made by a 33-person company in Santa Clara is listening. The platform watches air handlers and water loops, cross-references load patterns, and a few minutes later it suggests a setpoint change. The operator nods. Cooling energy drops. Over the next ninety days, it drops by eighteen percent.
The platform is called Atomiton. It does not advertise on stadium walls. Most engineers in the AI infrastructure conversation have never typed its name. And yet, somewhere between the carbon accounting startups and the legacy industrial control vendors, Atomiton has been quietly doing the unglamorous work of turning sustainability from a slide deck into a system that runs operations.
Atomiton calls it Immersive Intelligence. Strip the marketing and you get: a unified, real-time system that links what a factory or data center is doing to what it is consuming, and what it is emitting. The platform pulls energy, water, carbon and asset data into one place, models the resource-consuming processes as modular templates called BOTs, and runs the whole thing through a hybrid reasoning engine - deterministic rules where physics demands them, generative AI where pattern recognition pays off.
One real-time view of operations, resources and environmental impact across every facility.
Business Operations Templates - reusable blocks that model the parts of your plant that burn energy or move water.
Deterministic logic plus generative AI, producing recommendations an operator can defend in a control room.
Thing Query Language - because SQL was never written for chillers, turbines and tanks.
Live metrics, suggested actions, forecasted impact. The dashboard that argues with you.
Data centers, manufacturing, smart cities, oil and gas, precision agriculture, buildings.
Jane Ren spent years inside GE's digital business - one of the original team building what was then called the Industrial Internet. She watched the conversation about industrial software flip three times: from dashboards, to predictive maintenance, to ESG reporting. Each time, the same problem nagged her. The software always sat next to the operations, never inside them. Reports were generated. Slides were prepared. The actual machines kept running the way they had always run.
So in 2013, with Oleg Danilov and Alok Batra, Ren started Atomiton. The premise was almost dryly stubborn: if you want sustainability to matter, embed it in the daily decisions of the people who run the plant. Not in a quarterly memo. Not in an offset. In the setpoint.
The team built TQL, a query language whose entire reason for existing was that SQL had nothing useful to say about a heat exchanger. They built BOTs because every industrial customer wanted to start with templates and then customize. And they layered reasoning on top - first rule-based, then, as the AI wave arrived, a hybrid model that could weigh probability against physical constraint.
The result is a platform that, in 2026, sits in a particular sweet spot. Carbon accounting startups can tell you what you emitted last quarter. Atomiton tries to tell you what to change before the next quarter starts.
Figures self-reported by Atomiton customers via Energy Business Review APAC profile, 2025. Bars scaled for readability.
One of the founding team at GE Digital. Drives product and market direction at Atomiton; the public face of the company.
Co-founded Atomiton in 2013; engineering-side contributor to the original platform architecture.
Joined the founding team to help shape the company's industrial software bet.
Jane Ren, Oleg Danilov and Alok Batra start the company with a bet that industrial software needs to leave the dashboard and enter the loop.
Clearvision Ventures, BOLD Capital Partners and Envision Ventures back the company. Reported total raised around $6M.
CIOReview puts Atomiton on its list, citing the Digital Service Grid for Industrial IoT.
Platform extended to combine deterministic logic with generative AI for in-line recommendations.
Energy Business Review names Atomiton the exclusive recipient of the award, with a profile interview with CEO Jane Ren.
Most of the sustainability software market sells reporting. You count what you emitted. You publish a number. Atomiton sells the version of the problem that is harder to package and harder to demo, which may be why fewer companies do it: changing what the plant actually does, while it is doing it.
That distinction sounds academic until you read the customer numbers. An 18% cut in cooling energy at an AI data center is not a reporting artifact. It is a chiller running differently. A 12% improvement in fuel and electricity across a manufacturer with around a hundred sites is not an accounting choice. It is a hundred plants nudged in the same direction at the same time.
Run a data center cooler. Hit a corporate carbon target without burying it in offsets. Stitch a hundred manufacturing sites into one operations view. Catch a thermal imbalance before it becomes downtime. Argue with your own dashboard - and act on what it tells you.
TQL - Thing Query Language - exists because the people who built Atomiton decided SQL was the wrong shape for industrial things.
Business Operations Templates were called BOTs long before the abbreviation meant something else in the consumer AI world.
Roughly 33 people in Santa Clara are running the operations layer for customers with nearly a hundred sites.
Return to that hyperscale data center east of San Francisco. The Tuesday afternoon has become a Tuesday evening. Cooling energy is still tracking eighteen percent below where it sat in early spring. The operator who nodded at a setpoint suggestion three months ago is now reviewing forecasts on a dashboard that includes water consumption, renewable mix and projected carbon by week.
None of it is dramatic. None of it makes the press release rotation. The chillers hum the way chillers always have. But the building is doing something it could not do before: it is paying attention to itself, in the same units a CFO would care about, in the same minute the operations team is working. That is the thing Atomiton has been quietly trying to make normal for thirteen years. It looks, from a certain angle, like a small thing. Twelve percent here. Eighteen percent there. A hundred sites stitched together. A query language nobody asked for. And then, when you add it up across a customer base, it is the difference between a sustainability number on a slide and a sustainability number you can actually keep.
— END —