Applied Computing Raises $20M to Build an AI Brain for Oil and Gas Plants

Oil, gas, and petrochemical facilities generate mountains of data every second, yet most of it goes unused. A London-based startup called Applied Computing thinks it has found a way to change that, and investors are betting $20 million on the idea.

The company just closed a Series A round led by engineering giant KBR, with participation from Databricks Ventures, to build a foundation AI model built specifically for heavy industrial operations. The goal isn’t a chatbot or a coding assistant. It’s a model that can understand an entire refinery, gas plant, or petrochemical facility as a single connected system.

The Problem: Too Much Data, Too Little Insight

Founded in 2023, Applied Computing targets a specific and stubborn problem in heavy industry. A single oil, gas, or petrochemical facility can run thousands of sensors tracking everything from temperature and pressure to fluid velocity and viscosity. In theory, that’s a goldmine of operational data. In practice, most of it sits unused.

Co-founder and CEO Callum Adamson says facilities make operating decisions using less than 8% of the data available to them. Operators already collect the readings, he says, but they struggle to combine sensor data, engineering documentation, and the underlying physics and chemistry fast enough to turn it into a usable answer. Getting those three data sources talking to each other in real time is, in his words, the real challenge facilities face.

How Orbital Works

Applied Computing’s product, called Orbital, takes a different approach than typical AI models. Rather than functioning like a large language model that predicts the next word in a sentence, Orbital combines three types of models: a time series model, a physics-based model, and a language model, working together to predict the state of a facility.

The system analyzes live sensor readings while accounting for the physical and chemical constraints of the equipment involved, plus how operators are actually running the plant day to day. That combination lets engineers run simulations to see how a change in one part of a facility might ripple through the rest of its operations, before making the change in the real world.

The pitch to customers is speed. Applied Computing says Orbital can flag an anomaly, trace its root cause, and test whether a proposed fix might cause problems elsewhere in the facility, all within minutes. Adamson claims the platform can compress investigations that used to take days or weeks down to seconds, which in turn helps operators cut energy use while keeping output steady.

Early Traction and Customers

The startup says the response has been strong. It went from stealth mode to double-digit millions in annual recurring revenue in under 18 months, according to the company. Adamson says Orbital is already running at some large, publicly listed companies spanning upstream oil and gas, downstream refining, and petrochemicals, though he declined to say exactly how many customers the company has.

Its partner list includes Indian energy company Wipro and KBR, its lead investor, which has folded Orbital into its INSITE 3.0 digital platform for energy projects and is using the product for ammonia production. Adamson says the company is also working with a major U.S. upstream operator and plans to announce a partnership with a European oil major in the coming weeks.

A Crowded, Entrenched Market

Applied Computing isn’t entering an empty field. Established industrial software vendors already serve this market alongside newer AI-focused entrants. AspenTech sells simulation and AI-powered modeling software for upstream, refining, and chemical operations. AVEVA offers physics-based process simulation and “what-if” modeling for industrial plants. Cognite and Seeq focus more narrowly on the data layer, helping facilities analyze industrial data and build AI-driven workflows on top of it.

Adamson argues Applied Computing’s real advantage isn’t access to industrial data or process know-how, since larger incumbents already have plenty of both. Instead, he points to talent: attracting top AI researchers who want to work on frontier modeling problems rather than at a traditional energy company. He also points to the operational data Orbital collects through live deployments, which he says can’t be replicated with simulated data alone, since simulations can’t fully capture what happens inside a working plant.

The KBR partnership adds another layer to that advantage. Beyond capital, Adamson says it gives Applied Computing access to real operational data, deep industry expertise, and introductions to potential customers that would otherwise be hard to reach.

Expansion Plans

Applied Computing plans to put the new funding toward international expansion, hiring for research and engineering roles, and pursuing new deployments with energy clients. The company also announced a new office in Houston, adding to its headquarters in London and its operational hub in Bengaluru. Adamson says the U.S. base puts the company closer to two existing North American customers, and the company is also working on an expansion into the Middle East.

Why It Matters

Foundation models built for language have dominated the AI conversation for the past few years, but Applied Computing’s approach signals a different direction: models purpose-built around physics, chemistry, and industrial time-series data rather than text. For an industry where a wrong prediction can mean unplanned downtime, safety incidents, or wasted energy, that distinction matters. If Orbital can consistently turn sensor noise into fast, trustworthy answers, it could reshape how oil, gas, and petrochemical operators run some of the most complex and highest-stakes facilities in the world.