foundation models

A continuously evolving digital model of the asset

Beneath the application products sits a layer of foundational domain models. They are linked to one another and evolve across the entire well and field life cycle — so an answer about cost is also an answer about the subsurface, the rig and the pipe.

Why tde one can build them

Foundation models of an industrial asset are not a modelling problem so much as a data problem. tde one is uniquely positioned to build them: they are created by combining deep domain knowledge with the proprietary data already accumulated across 70,000+ well construction projects and 1,800 evaluated rigs — an advantage no generic AI platform can replicate.

tde one assistant

Question: What is the expected well cost if we switch to a 1,500 HP rig and modify the casing design?

Answer: Based on the current earth model, well model, process model, resource model and facilities model, estimated well cost decreases by 6.4% while maintaining drilling objectives.

Five linked models

tde earth model

Subsurface intelligence

  • Seismic
  • LWD / MWD
  • Petrophysics
  • Reservoir understanding

Continuously updated geological and geomechanical representation of the subsurface.

tde well model

Well intelligence

  • Well design
  • Construction
  • Production
  • Plug & abandonment (P&A)

Single digital representation of the physical well across its entire life cycle.

tde process model

Operational intelligence

  • Operational sequences
  • Best practices
  • Execution workflows
  • Continuous learning

The operational language that connects planning, execution and performance.

tde resource model

Resource & cost intelligence

  • Rig
  • Drill string
  • Materials
  • People

Complete representation of resources and full life-cycle cost of a well.

tde facilities model

Facilities intelligence

  • Wellheads
  • Flowlines
  • Separators
  • Gathering systems
  • Export pipelines

Digital representation of the surface production network linking wells, processes, resources and cost across the full field life cycle.

A producing field seen from the air under flat grey light: wellheads and gathering lines laid out in a grid between separator skids, storage tanks and access roads, running to a distant horizon.
ai reasoning across foundational models

Five models, one line of reasoning

The models are queried together, not one at a time. That is what lets a question about the rig come back as an answer about cost.

  • Connected intelligence

    Reasoning spans the full asset, not one model at a time.

  • Continuously updated

    The models change through operations, as the operations happen.

  • Proprietary data

    Built on tde's own accumulated data and domain expertise.

  • Purpose-built

    For the energy industry, rather than adapted to it.

powered by unique domain knowledge

What the models are trained on

70,000+
well construction projects
1,800
evaluated rigs
Deep domain expertise
Built over decades in the energy industry
Continuous learning
Models evolve with every operation
Proprietary & secure
Your data. Your models. Your advantage.

One platform. All operations. Smarter together.

tde ingest feeds tde pronova, tde pronova feeds the foundation models, the models feed the application products, and tde one is what you actually talk to. Five layers, one answer.

talk to us about your operations