Limited published evidence · Oil and gas
AI well report extraction: Petoro and SLB example
Petoro and SLB describe Well Schematic Extractor and Well History applications for well planning. LLM-based agentic workflows extract engineering information from reports and logs, with deterministic checks and visual quality control. The source reports only preliminary user testing.
Well-report extraction and well-history preparation
Petoro and SLB
Implemented applications with preliminary user testing reported; broad production rollout is not established.
A published third-party example, not a Nuvepro customer case or evidence of a partnership.
The workflow at a glance
Petoro and SLB reported work
Illustrative summary of the published account
Reported work components. Their arrangement does not imply a fixed execution order or simultaneous work.
- Engineering applications
Well Schematic Extractor and Well History
Implemented tools support well planning with preliminary user testing.
- Document work
Agentic engineering extraction
Reports and logs provide well information.
- Quality mechanism
Deterministic checks and visual review
The source describes checking extracted representations.
What the source reports
SLB describes Well Schematic Extractor and Well History applications for well planning. LLM-based agentic workflows extract engineering information from reports and logs, with deterministic checks and visual quality control.
Reported technology: SLB; underlying model provider not specified. Scope and outcomes are source-reported.
What the enterprise environment would need
The following is a proposed project design for an enterprise doing similar work. Published details are marked; the remaining systems, data and scenarios would be agreed for your project.
Proposed practice environment
Reconcile a well-history evidence package
Illustrative project design to scope with your enterprise.
- 1 · Set up the environment
Inconsistent well reports
Fictional well IDs, revisions, depths, casing/completion fields and approved reference schematics.
- 2 · Preparation and agent contributions
Prepare well-history package
Extract values and source locations, retaining units and conflicting references.
- 3 · Independent checks
Independent engineering reconciliation
Check depths, units and provenance against specialist-approved schematics.
- 4 · People decide acceptance
Engineering specialist accepts evidence
Review unresolved discrepancies before accepting the bounded extraction version.
A reviewable evidence package
A traceable schematic, well-history summary and discrepancy register.
If a check fails → withhold acceptance.
Incompatible depth references are merged into a confident but wrong schematic. Withhold acceptance, repair, independently recheck and return to owner review.
Failure or owner rejection returns to review
- Hold
- → 2 · Revise work
- → 3 · Rerun independent checks
- → 4 · Owner review
Operating boundary. Controlled document tools only; no well-plan approval, field control or autonomous drilling.
- Systems and tools
- Proposed: Use a controlled well-document workspace, extraction tools and an engineering review interface.
- Data and state
- Proposed: Preserve well identity, document revisions, casing/completion information, units and source locations.
- People and handoffs
- Proposed: Drilling or reservoir specialists validate the extracted representation before planning use.
- Rules and authority
- Proposed: Keep extraction separate from authority to approve a well plan or operate field equipment.
- Normal and exception cases
- Proposed: Include illegible logs, inconsistent depths, mixed units and conflicting historical documents.
- Reset and replay
- Proposed: Freeze document bundles and reference schematics so revised extraction can be compared.
A proposed simulation
An illustrative practice design derived from this work, not an executed assessment.
Extract a well-history package from inconsistent reports and prepare a traceable schematic for engineering review.
Review gate: Specialists can reconcile material values, units and provenance, with unresolved discrepancies explicitly held for review.
Challenge the result: Supply two incompatible depth references; a confident merged schematic must fail the engineering check.
What needs checking
- Proposed: reconcile extracted fields against engineer-approved schematics and source references.
- Proposed: test depth/unit consistency and missing-information escalation independently of a fluent well-history summary.
Review environment fidelity, human capability and agent performance separately. The enterprise’s accountable owners decide acceptance of the shared workflow and any operating use.
Questions about this example
- What do Petoro and SLB use AI for?
- SLB describes Well Schematic Extractor and Well History applications for well planning. LLM-based agentic workflows extract engineering information from well reports and logs, and the source says deterministic checks and visual quality control are applied.
- Who checks the extracted well information?
- The source mentions deterministic checks and visual quality control. It does not name who reviews the results. A proposed design, not from the source, has drilling or reservoir specialists validate the extracted representation before planning use.
- How strong is the evidence for this well-planning example?
- The evidence is limited and early-stage. The applications are implemented and preliminary user testing is reported, but user count, test duration, extraction accuracy and production breadth are not supplied. It is not proof of autonomous drilling or operating approval.
- What would an enterprise need to build something similar?
- This is Nuvepro's proposed design, not Petoro or SLB's. It would need a controlled well-document workspace, frozen document bundles and engineer-approved reference schematics. Tests would include illegible logs, inconsistent depths and mixed units, checked against source references.
Shape this around your enterprise
Bring a similar workflow and the people responsible for it. Together, define the records and systems to use, the result your team will build or configure, and how its owners will check it. Nuvepro scopes the simulation, environment and preparation. Agree any existing-agent evaluation, implementation assistance and production responsibilities separately.
Sources and limitations
Curation review: . This is the date this source example was reviewed, separate from source publication dates.
- User count, test duration, extraction accuracy and production deployment breadth are not supplied.
- The page displays 06/05/2026; its numeric date convention was not resolved, so no exact publication date is asserted.
- This is an early or limited-evidence example, not proof of autonomous drilling, field control or operating approval.