Applied Computing raises $20M for a plant-wide oil and gas AI model
A London startup with a Bengaluru engineering hub just raised $20M to build a foundation model that reads an entire refinery in real time. India's refiners should be watching closely.
The News
Applied Computing, a London-based industrial software startup founded in 2023, has raised a $20 million Series A to build what it calls a foundation AI model for oil, gas, refining and petrochemical plants. The round was led by the engineering giant KBR, with participation from Databricks Ventures.
The company's product, Orbital, is designed to model an entire facility rather than a single valve or vessel. It fuses three ingredients that rarely sit together: time series analysis of live sensor feeds, physics-based modelling of the underlying chemistry, and language processing over engineering documentation. The result, the company says, is a system that can predict how a plant will behave, flag anomalies, and simulate operational changes across connected units.
Applied Computing runs its headquarters in London, an engineering hub in Bengaluru, and a newly opened office in Houston. Chief executive and co-founder Callum Adamson framed the opportunity bluntly: "It's an AI problem. It's not a data problem, and it's not an energy problem." The startup says it has already reached double-digit millions in annual recurring revenue within 18 months, selling to large, publicly listed upstream and downstream operators.
Why It Matters
Heavy industry has spent two decades wiring up its plants and then largely ignoring the readings. Applied Computing claims operators currently use less than 8% of the sensor data they collect. That gap is the pitch: a model that turns dormant telemetry into decisions, compressing investigations that once took days or weeks into seconds.
The interesting signal here is who is writing the cheque. KBR is not a venture tourist; it builds and maintains the physical plants this software is meant to watch, and has folded Orbital into its INSITE 3.0 platform. When an incumbent funds a challenger rather than build in-house, it usually means it has concluded the hard part is the model, not the distribution.
Applied Computing is also walking into a crowded room. AspenTech, AVEVA, Cognite and Seeq have all sold industrial analytics for years. The bet is that one foundation model, trained to reason across a whole facility, beats a stack of narrow point tools.
Indian Angle
India is one of the most concentrated bets in the world on exactly the assets Orbital is built for. Reliance Industries' Jamnagar complex is the single largest refining site on the planet, and public-sector heavyweights such as Indian Oil, BPCL, HPCL, ONGC and GAIL operate sprawling refining and petrochemical estates that generate precisely the under-used sensor data Applied Computing wants to mine. A tool that squeezes more throughput and fewer unplanned shutdowns out of existing hardware maps directly onto India's refining margins.
The India connection is not hypothetical. Applied Computing already runs its engineering hub in Bengaluru and counts Wipro among its integration partners, which means Indian talent and an Indian IT services major are both wired into the roadmap. For Wipro and its peers, embedding a foundation model into brownfield energy plants is exactly the higher-margin work they have chased as classic outsourcing commoditises.
There is a safety dimension too. As Indian refiners lean on AI to run closer to operational limits, the Petroleum and Explosives Safety Organisation and the sector's own process-safety norms will need a view on model-driven decisions in hazardous facilities. Indian enterprises tend to adopt fastest when a global systems integrator carries the deployment risk, and a KBR-plus-Wipro pairing fits that pattern neatly.
FAQ
How much did Applied Computing raise and who led the round?
The company raised a $20 million Series A led by KBR, with participation from Databricks Ventures. It was founded in 2023 and is headquartered in London with hubs in Bengaluru and Houston.
What does the Orbital model actually do?
Orbital is a foundation model for process plants. It combines live sensor time series, physics-based modelling and language processing over technical documents to predict facility behaviour, flag anomalies and simulate operational changes across an entire site rather than one component.
Who are its main competitors?
Applied Computing competes with established industrial analytics vendors including AspenTech, AVEVA, Cognite and Seeq. Its differentiation is a single whole-plant model versus a collection of narrower point tools.
Why should Indian readers care?
India hosts the world's largest refinery at Jamnagar and vast public-sector petrochemical assets. Software that lifts throughput and cuts downtime targets India's refining economics directly, and the startup already runs a Bengaluru hub and partners with Wipro.
This story was reported by TechCrunch. Read the full original coverage at TechCrunch.