Quantemplate announces breakthrough in AI-driven data engineering, enabling machines to suggest commercial data pipelines from data semantics and Network Intelligence

UK-based Quantemplate develops new AI infrastructure that can identify the meaning of unfamiliar commercial data, determine how it should be transformed, and recommend the workflow required to make it production-ready.

‍
London, UK – 21 September 2026 – Quantemplate, the peer-to-peer network for commercial data communication and automation, today announces a significant development in its AI. The new technology analyses the semantics of incoming data and combines that understanding with intelligence from its commercial data network to suggest how data-processing pipelines should be built.

The development moves beyond AI-assisted coding or conventional data mapping. Rather than requiring a forward deployed engineer or implementation consultant to determine how unfamiliar datasets should be connected, cleansed, mapped and transformed, Quantemplate’s technology analyses the structure and meaning of the data and uses knowledge embedded in the network’s social graph to determine how the resulting workflow should be constructed.

In a demonstration of the technology published on the Quantemplate YouTube channel, raw datasets can be converted into a production workflow in minutes. The system identifies the characteristics and semantics of columns, determines relationships between datasets and recommends the data dictionaries, column mappings and join points required to produce usable data.
The objective is to make sophisticated data engineering accessible to all businesses without the traditional cost and skill constraints inherent with teams of implementation consultants, systems integrators or forward-deployed engineers.

‍

A different approach to enterprise AI

Quantemplate believes that a fundamental limitation of today’s AI systems is that intelligence is often separated from the data infrastructure on which businesses operate.

Large Language Models (LLMs) can generate code and answer questions, but they do not inherently know how the commercial datasets of thousands of organisations relate to one another, which fields represent equivalent concepts, which transformations have worked previously, or which workflow is appropriate for a particular business process.
Quantemplate’s architecture is designed to address this problem by combining semantic understanding of data with Network Intelligence.

Organisations using the network can opt-in to Network Intelligence, which surfaces relationships between data sources, counterparties, datasets, transformations, mappings, workflows and business processes. Those relationships form a social graph from which the platform can derive recommendations for other participants, while individual organisations retain control over their own data.

The result is intended to create a form of industry-wide neural network for commercial data: intelligence generated through the interactions of many organisations can improve the ability of the system to understand and process data for each participant.
Quantemplate already operates as a peer-to-peer commercial data network, connecting organisations across value chains and supporting the exchange, reconciliation, transformation and activation of
business data.

‍

Twenty years of development behind the technology

Quantemplate’s technology has been developed over two decades, with its origins in the insurance and financial-services industries, where the company has worked on some of the most data-intensive commercial processes.

The platform has been used operationally in the insurance sector since 2013. According to Quantemplate’s internal records, its technology has processed data associated with approximately $67 billion of combined customer revenue, receiving data from 14,188 trading partners across 75 countries.

The data entering the network spans industries including capital markets, finance and insurance, energy, healthcare, pensions, automotive, aviation, marine, construction, real estate, retail and professional services.

Learning from the natural world

The architecture behind Quantemplate was inspired partly by one of nature’s most successful distributed networks: the mycorrhizal network, sometimes referred to as the ‘Wood Wide Web’. In a forest, microscopic fungal networks connect the root systems of different plants and trees, allowing resources and biochemical signals to move through the ecosystem. Individual organisms remain autonomous while the network enables information and resources to flow between them.
Quantemplate applies a similar principle to commercial data by providing the connective infrastructure through which organisations can exchange data and develop relationships with other participants. The network therefore becomes more useful as more organisations participate: each new relationship provides additional context for understanding data, while the underlying data remains under the control of its owner.

A capital-efficient AI Lab

Quantemplate has been cash-flow break-even since the 2020 management buyout, with software development funded through customer licence fees and the HMRC’s R&D tax credits scheme.

Founders and employees collectively own approximately 92.5% of Quantemplate which has enabled Quantemplate to remain focused on developing this highly complex technology over a prolonged development timeframe. The company’s latest AI development represents the culmination of this long-term approach: combining proprietary semantic understanding of commercial data with the Network Intelligence generated by relationships between organisations, datasets and workflows.

A million rows for free: bringing smaller businesses into the network

Quantemplate has developed an economic model intended to make advanced data infrastructure accessible beyond large enterprises. The platform is available to organisations at no cost for up to one million rows of generated data per month, with unlimited data licences starting at $499 per month.

From a technical perspective, the network is served by a Scala-centric backend, a software choice that reflects both performance and approach. Scala provides exceptional robustness and scalability, while its open-source nature allows continuous improvement through community contribution.

Running on commoditised CPUs via the JVM (rather than GPU’s favoured by LLM’s), it also enables the platform to operate at extremely low cost, making the network freely accessible to all organisations. The company uses surplus income from enterprise licences to subsidise compute costs for smaller participants in the network. The model is deliberately analogous to an ecosystem: larger enterprises provide the economic energy that allows smaller businesses to participate in the network, increasing the overall breadth and utility of the system. Even the largest businesses depend on their trading partners, making shared data infrastructure vital despite competitive rivalries – much as forest networks connect established trees and saplings, sustaining the wider ecosystem.

A commercial network already forming

The network currently has registered organisations collectively representing approximately $730 billion of revenue and $6.8 trillion of assets, according to Quantemplate. The company says these organisations represent approximately 0.66% of global GDP and 1.4% of global financial assets. Quantemplate’s ambition is to create an open commercial data network available to organisations of any size and in any industry.

A new category of AI infrastructure

Quantemplate’s technology sits at the intersection of three developments in enterprise technology: artificial intelligence, increasingly distributed commercial data and the automation of business workflows.

The company’s proposition is that a critical infrastructure required for the successful deployment of predictive AI models in business is the automated supply of accurate and clean data. This requires an infrastructure that can understand the meaning, relationships and provenance of commercial data itself.
Its technology combines four elements:
‍
Semantic AI that identifies the meaning and characteristics of unfamiliar data.

Network Intelligence derived from relationships between organisations, datasets, workflows and business processes.

Automated data pipelines that translate those insights into executable production workflows.

Continuous data-quality improvement enabling errors and missing information to be identified and corrected as data moves through the network.
‍
The company believes this combination could materially reduce the human effort traditionally required to gather, prepare, route and analyse commercial data.

Research & Development Team

Adrian Rands, co-founder, leads product architecture and roadmap. Before Quantemplate, Rands worked as a reinsurance broker at Howden and as Entrepreneur in Residence at hedge fund D.E. Shaw.

Tom de Gay, co-founder and Head of Design, leads product design, customer engagement and the company’s Help Centre, website and video content.

The development team includes specialists in Scala and PostgreSQL for the backend and React for the frontend.

“The system is beginning to understand the data”

“Data is becoming the fundamental infrastructure of the modern economy, but most businesses still have to employ teams of people to explain what their data means and how it should be connected,” said Adrian Rands, co-founder of Quantemplate.

“We have spent twenty years building the underlying data infrastructure. The breakthrough is that the system can now understand the semantics of unfamiliar commercial data and use the intelligence of the network to determine how and where that data should be used.”

‍
“That changes the role of the data engineer. Instead of manually constructing every pipeline, the human increasingly becomes the person who supervises, approves and corrects what the AI has understood and proposed.”

The next generation of enterprise data infrastructure

Quantemplate’s thesis is that the next generation of AI infrastructure will need to understand not just language, but data, relationships and business processes.

If AI can understand what a column of data means, identify equivalent information elsewhere in the network, understand how that information has previously been transformed and then construct the required workflow, the cost and complexity of enterprise data engineering could change fundamentally. Quantemplate is now seeking to extend its network across industries and geographies, with the objective of making advanced data infrastructure accessible to organisations that historically could not afford enterprise data engineering systems.

About Quantemplate

Quantemplate is a peer-to-peer commercial data communications and automation network. The platform enables organisations to exchange, reconcile, cleanse, transform, validate and orchestrate business data and workflows.

The company describes its technology as a social network for commercial data, combining semantic AI, data-quality automation, workflow orchestration and Network Intelligence.

Data remains under the control of the organisation that owns it, with access granted to authorised parties.
‍

Contact

Adrian Rands
adrian.rands@quantemplate.com
+44 (0) 7940 531 454
www.quantemplate.com

‍