Ontology & Semantic Knowledge Graphs Strategic Positioning | TotalEnergies x Forvis Mazars

Tool Benchmark

How do we make a semantic tooling benchmark credible, comparable, and procurement-ready?

Objective: Evaluate SousLeSens against alternatives with transparent criteria for procurement support.

×

Contents

Tool Benchmark

1.Context

2.Executive summary

3.Scoring framework

4.Implications for TotalEnergies

5.Architecture boundaries

6.SousLeSens

7.Microsoft Fabric

8.Atlas IA (AtlasAI)

9.RDFox

10.Metafactory

11.Collibra

12.Cognite CDF

13.Databricks OntoBricks

14.Palantir Foundry / AIP

15.Conclusion

Context

Ontology & Semantic Knowledge Graphs Strategic Positioning

TotalEnergies and OneTech already have a solid semantic foundation through TSF and SousLeSens (SLS). The strategic issue is adoption: leadership needs to see why governed meaning matters, where it creates measurable value, and how it fits the OneTech landscape across platforms, operations, data products, and AI.

The work builds on a first mission conducted from November 2025 to January 2026, the Cross-FPSO Semantic Preparation and Ontology Vetting Program, where Forvis Mazars assembled the team with NCOR, the National Center for Ontological Research. That mission validated TSF's alignment with BFO (Basic Formal Ontology, ISO/IEC 21838-2:2021) through a formal audit of the TE Business Objects ontology, and delivered a machine learning prototype that matched 23,000 Pazflor functional locations to Dalia equivalents to bootstrap cross-FPSO job card transfer for life extension planning. This second mission turns that validated technical foundation into a positioning narrative for TSF and SousLeSens within OneTech.

5 coordinated workstreams and 5 client-facing decks.
12 internal stakeholder interview profiles and 3 Voice of Peers interview profiles.
9 benchmarked tool profiles across semantic control, platform, graph, catalog, and AI layers.
25 strategic intelligence organization profiles backed by 26 evidence notes.
Management assets include the Executive Deck, one-pager, FAQ, use-case showcase, decision framework, RFI/RFP support, benchmark matrix, and evidence registers.

This deck is the platform and procurement evidence layer. Use it to separate semantic authority from storage, workflow, analytics, catalog, and AI consumption, and to decide what each tool should consume, govern, or prove before OneTech treats it as part of the semantic architecture.

Executive summary

Tool benchmarking should provide objective procurement support for selecting the right semantic platform role.

How do we make a semantic tooling benchmark credible, comparable, and procurement-ready?

Target scope

Proposal tools

SousLeSens, Microsoft Fabric, Atlas IA (AtlasAI), RDFox, Metafactory, Collibra, and Cognite CDF are evaluated as the core proposal scope. Databricks OntoBricks and Palantir are added as comparators.

Methodology

Nine criteria, two axes

Nine weighted criteria roll up to semantic authority (Y) and operational power (X) for quadrant placement, plus an overall procurement score. See the scoring framework slide for definitions.

Deliverables

Procurement support pack

Comparative matrix with quantitative scores and documented rationale.
Technical evaluation report and recommendation summary.
RFI/RFP question bank for vendor evidence.

Governance

Transparent scoring, documented rationale, vendor evidence traceability, and final decision authority retained by TotalEnergies.

Scoring framework

Nine weighted criteria define semantic authority, operational power, and the overall benchmark score.

What does each criterion measure, and how does it drive placement on the quadrant?

Y: Semantic authority

Vertical axis on the positioning chart. Can the tool own, govern, and prove meaning?

Open standards & portability (weight 15)
Semantic governance (weight 12.5)
Reasoning & validation (weight 15)

Y = (OSP×15 + SG×12.5 + RVE×15) ÷ 42.5

X: Operational power

Horizontal axis on the positioning chart. Can the tool integrate, deliver AI, and get adopted at scale?

Integration & industrialization (weight 12.5)
AI & agents (weight 10)
UX & adoption (weight 5)

X = (II×12.5 + AIA×10 + UEA×5) ÷ 27.5

Procurement context

Three additional criteria inform the overall weighted score in the comparative matrix. They do not set quadrant position directly.

Solution overview (weight 5)
Architecture layer fit (weight 15)
Total cost of ownership (weight 10)

Reading the chart

Each vendor is scored 1 (limited) to 5 (proven). Dashed lines on the quadrant sit at 3.0 on each axis. Tools above 3.0 on Y carry stronger semantic authority; tools above 3.0 on X carry stronger operational power.

Implications for TotalEnergies

The SKG target pairs semantic authority with operational power; most vendors lead on only one axis.

Where should TotalEnergies anchor meaning, and which tools can execute delivery without diluting tradecraft?

SousLeSens: reference platform Proposal scope Comparator Dashed lines at score 3.0 on each axis.

Why each tool sits here

SousLeSens (4.35 / 3.45) sits near the industrial semantic platform because TE already runs OWL/RDF tradecraft, governance, and reasoning through TSF; it does not sit further right because group-wide operational delivery and adoption still need to scale.
Metafactory (4.65 / 3.64) scores at the top as the strongest external ontology workbench comparator; it is benchmarked for parity on ontology management, not proposed as TE's semantic control layer.
RDFox (4.41 / 2.91) lands in semantic authority because its value is standards-based reasoning and validation; it is positioned left of center because it is a reasoning component, not an industrial delivery platform.
OntoBricks (2.55 / 2.65) sits in emerging because Databricks Labs is still maturing the link from Unity Catalog to ontology graphs; it enters the benchmark when the lakehouse is in scope, not as a governed enterprise semantic backbone.
Atlas IA (AtlasAI) (2.35 / 2.25) sits in emerging as a young knowledge-graph vendor without yet proving open-standards depth, industrial references, or TE-scale deployment; RFI must confirm product scope and fit.
Microsoft Fabric (2.20 / 2.40) sits in emerging on this chart because its strength is enterprise data and analytics consumption, not authoritative ontology tradecraft; TE should use it downstream of TSF, not as the meaning layer.
Collibra (2.88 / 3.82) sits on the operational side because catalog, stewardship, and policy scale in the enterprise; semantic authority stays moderate because it governs metadata, not formal disambiguation or reasoning.
Cognite CDF (2.65 / 4.45) sits far right as an industrial operational platform for context, integration, and data ops; it should consume TSF definitions rather than replace the semantic control layer.
Palantir (2.94 / 5.00) sits at the operational edge as the market comparator for ontology-enabled operations and AI actioning; TE keeps portable meaning in TSF / SousLeSens because Palantir optimizes execution inside a proprietary stack.

Architecture boundaries

Use the benchmark to test architecture boundaries and evidence quality.

What should the benchmark contribute to the communication kit?

Boundary 1

Semantic ownership

SousLeSens / TSF should govern definitions, rules, standards alignment, and semantic model portability.

Boundary 2

Reasoning and management

RDFox, Metafactory, and OntoBricks should be tested for reasoning depth, ontology lifecycle, lakehouse mapping, and standards fit.

Boundary 3

Platform consumption

Fabric, Collibra, Cognite CDF, Databricks, Palantir, and Atlas IA (AtlasAI) should consume governed meaning where it improves decision support.

Communication position

The benchmark should show where each tool belongs, what TotalEnergies must govern itself, and which vendor claims require proof.

Tool profile

SousLeSens

Open-source ontology workbench and TotalEnergies semantic control layer

Signal

Best fit for controlled definitions, semantic rules, standards alignment, model portability, and cross-platform reuse.

Main processes

Ontology design and concept lifecycle.
RDF / OWL / SPARQL / SHACL modeling.
BFO-aligned semantic modeling and validation.
Mappings from source systems to semantic models.

Traceability

Model equipment families and tag variation across assets.
Represent failure modes, inspections, risks, and evidence.
Publish governed concepts to consuming platforms.
Make decision logic auditable and reusable.

Implications

Position SousLeSens as the semantic control layer; strengthen operating model, support, and integration with platform teams.

Tool profile

Microsoft Fabric

Microsoft analytics, BI, data platform, and Copilot ecosystem layer

Signal

Strong adoption fit for reporting and Microsoft ecosystem consumption; formal ontology depth and model portability need RFI proof.

Main processes

Power BI semantic models and reporting.
Lakehouse and data engineering in Microsoft stack.
Copilot and AI-assisted consumption.
Prepared analytical datasets for business users.

Traceability

Expose semantic outputs to management dashboards.
Provide governed metrics and reporting views.
Consume governed concepts for consistent labels and definitions.
Clarify graph model limits and export capability.

Implications

Use Fabric as a consumption and reporting layer; keep formal ontology ownership and cross-platform meaning in SousLeSens / TSF.

Tool profile

Atlas IA (AtlasAI)

Private legal AI knowledge graph platform for firm-owned intelligence

Signal

Relevant as a knowledge graph platform focused on document, entity, relationship, and institutional-memory intelligence.

Main processes

Document and system ingestion.
Entity and relationship extraction.
Firm-owned knowledge graph construction.
Cited answers, search, drafting, and workflow support.

Traceability

Index technical documents, procedures, reports, and precedent material.
Map entities such as equipment, contracts, reports, parties, and dates.
Answer questions with links back to source evidence.
Test whether AtlasAI can align extracted entities with governed concepts.

Implications

Keep Atlas IA in scope as a knowledge graph candidate; confirm standards support, enterprise fit, and industrial-domain adaptability during RFI.

Tool profile

RDFox

High-performance RDF knowledge graph and reasoning engine

Signal

Strongest fit for reasoning depth, inference, rules, and explainability; it works as an engine connected to ontology and governance tooling.

Main processes

RDF graph storage and SPARQL querying.
OWL 2 RL ontological reasoning.
Datalog rules with aggregation and negation.
Materialization and incremental reasoning.

Traceability

Infer similar equipment patterns from formal rules.
Validate mission-critical decision logic against explicit constraints.
Precompute derived facts for fast decision queries.
Provide explainable rule paths for AI or BI consumers.

Implications

Evaluate RDFox as a reasoning component connected to TSF / SousLeSens and governed through ontology management workflows.

Tool profile

Metafactory

Enterprise knowledge graph platform for semantic modeling and ontology governance

Signal

Strong fit for ontology management, visual modeling, collaboration, and semantic model publication based on W3C standards.

Main processes

Visual ontology and vocabulary management.
OWL / SHACL / SKOS-oriented modeling.
Ontology lifecycle, review, and publication.
Knowledge graph applications and discovery.

Traceability

Compare with SousLeSens on visual modeling and governance workflow.
Test BFO and industrial ontology compatibility.
Evaluate model publication and reuse by business users.
Assess integration with catalogs and data platforms.

Implications

Use Metafactory as the most direct external comparator for SousLeSens on ontology management and business-user adoption.

Tool profile

Collibra

Enterprise data catalog, ownership, lineage, and semantic layer integration

Signal

Strong complement to TSF for ownership, metadata, stewardship, semantic layer mapping, and governance workflows around data assets.

Main processes

Catalog, glossary, lineage, and metadata governance.
Dataset ownership and stewardship workflow.
Semantic layer mapping to business assets.
Audit, approval, and governance evidence.

Traceability

Answer who owns this dataset and whether it can be used for reporting.
Link governed datasets to semantic concepts.
Track lineage between source, semantic model, and reporting output.
Support data governance roles around operational evidence.

Implications

Use Collibra for stewardship and evidence of governance; connect catalog terms to TSF and avoid parallel definitions.

Tool profile

Cognite CDF

Industrial DataOps platform for contextualized operational data

Signal

Strong fit for operational navigation and contextualization across assets, time series, engineering documents, events, and field data.

Main processes

Asset hierarchy and equipment context.
Time series, historian, and operational source integration.
Entity matching and contextualization workflows.
Industrial knowledge graph and digital twin support.

Traceability

Find similar equipment across industrial assets.
Connect anomalies to maintenance and inspection evidence.
Expose operational context for mission-critical decisions.
Support fast navigation by engineering and operations teams.

Implications

Use Cognite CDF for operational context while TSF governs definitions, similarity logic, and semantic rules consumed by the platform.

Tool profile

Databricks OntoBricks

Knowledge graph builder for Databricks lakehouse data

Signal

Promising bridge between lakehouse data and semantic standards; industrial maturity and governance fit still require RFI proof.

Main processes

Unity Catalog metadata import.
OWL ontology design and industry-standard ontology import.
R2RML mapping and triple materialization.
OWL 2 RL, SWRL, SHACL reasoning, GraphQL, and MCP access.

Traceability

Map lakehouse tables to governed semantic concepts.
Materialize triples from operational and reporting data.
Run semantic reasoning and quality checks in the Databricks environment.
Expose graph outputs to APIs, agents, and analytical consumers.

Implications

Evaluate OntoBricks when Databricks is in scope; keep TSF / SousLeSens as the governed semantic authority consumed by the lakehouse.

Tool profile

Palantir Foundry / AIP

Ontology-driven operational decision and AI action platform

Signal

Strong benchmark for operational ontology, decision workflows, and AI actioning; open standards, portability, and lock-in require sharp RFI scrutiny.

Main processes

Ontology object, link, property, and action modeling.
Operational applications built on governed enterprise context.
AIP agents, automations, and human-in-the-loop workflows.
Security, permissions, writeback, and operational control.

Traceability

Represent operational objects, events, evidence, and decision actions.
Connect recommendations to governed operational workflows.
Trigger approved actions from AI-assisted decision paths.
Benchmark against the semantic framework for ownership and portability.

Implications

Use Palantir as the market comparator for ontology-enabled operations; keep TSF governance explicit so TotalEnergies owns meaning across platforms.

Conclusion

No single platform governs meaning and reasoning together: TSF and SousLeSens are the only credible candidates for that layer.

Nine tool profiles across five architectural layers confirm the central finding: no single platform governs formal meaning, runs reasoning, manages operational context, and exposes it to AI at the same time. The most durable architecture keeps semantic authority separate from execution. SousLeSens and TSF are the clearest fit for that authority role, with BFO alignment, W3C standards, model portability, and formal governance built in.

The benchmark also clarifies the positioning logic. Competing with Cognite, Collibra, or Fabric on their own ground is the wrong frame. Each platform serves a defined layer: operational context, stewardship and lineage, reporting and analytics. TSF governs the meaning those platforms consume. The value of that separation grows as the number of consuming platforms and AI agents increases, because governed definitions travel across vendor boundaries while local semantics do not.

SousLeSens is the most standards-aligned option for formal semantic control and ontology governance.
Metafactory is the closest direct comparator: an RFI should test BFO compatibility and governance workflow depth.
RDFox brings the strongest reasoning capability and is best evaluated as a component connected to TSF.
Collibra complements TSF on ownership and lineage; the boundary between catalog terms and ontology concepts must be explicit.
Palantir is the strongest benchmark for ontology-enabled operations and AI actioning; open standards and portability require scrutiny.

The procurement process should use this benchmark as the starting structure, with the RFI/RFP question bank validating vendor claims before any platform selection decision is made.