Cognite Data Fusion
Alternative for EPC Delivery, Not Just Operations

Operations. Cognite is built for owner-operators who already have a commissioned plant with extensive OT/IT data streams and want to build a knowledge graph over all of it for operational use cases (predictive maintenance, root cause analysis, operational agents). It is not designed for EPC project delivery deliverables.

Not natively. Cognite contextualises tags that are already in structured form: it links them to asset hierarchies. For legacy scanned or PDF P&IDs typical of brownfield plants and Indian / Gulf mid-EPC projects, an upstream extraction step is required. Pathnovo is the natural upstream partner: Pathnovo extracts, Cognite contextualises.

For Indian PSU refineries (IOCL, BPCL, HPCL, ONGC) with budgets and multi-year operational transformation programmes, Cognite is a serious option: Saudi Aramco's 7.4% stake signals the commitment. For Indian mid-EPC contractors (Tata Projects, Toyo India, KPIL, KBR India) delivering projects rather than operating plants, Cognite is over-scoped and over-priced. Pathnovo is the better fit.

Yes. They are complementary, not competitive. Pathnovo extracts structured engineering data from project documents. Cognite contextualises that data into a full operational knowledge graph. At owner-operator scale, using Pathnovo upstream (EPC extraction) and Cognite downstream (operations) is an established pattern.

Cognite Data Fusion implementations are estimated at $500K+ per year with 12–24 months of integration work. Pathnovo is credit-based, full platform included on every paid tier, free 14-day trial, and Enterprise contracts for multi-asset operators. For an EPC project delivery use case, the commercial models are not comparable: they serve different buyers. See pricing for plans.

Cognite Data Fusion and GE Digital Predix (now part of GE Vernova) are both industrial platforms but target different problems. Predix originated in 2015 as an industrial IoT platform for asset performance management on rotating equipment and power generation fleets, time-series-first, optimised for sensor data, vibration analysis, and predictive maintenance on GE-installed assets. Cognite Data Fusion is contextualisation-first, ingesting heterogeneous OT/IT data (P&IDs, asset hierarchies, sensor streams, ERP records, documents) into a unified knowledge graph for cross-domain operational use cases. In short: Predix is APM for industrial fleets, Cognite is the operational backbone for owner-operators. Where they overlap is predictive maintenance and asset health, where Cognite has caught up and Predix's market share has compressed since the original 2015 vision narrowed under GE Vernova. Pathnovo does neither of these jobs: Pathnovo is upstream of both, extracting structured engineering data from PDFs and scanned drawings that either platform then ingests.

Three different jobs. Pick GE Digital Predix if your scope is fleet-wide asset performance management on GE-installed rotating equipment, gas turbines, generators, and aero-derivative assets where Predix's time-series APM models and OEM data access are mature. Pick Cognite Data Fusion if your scope is whole-operations contextualisation across OT and IT, multi-vendor assets, and you want a single knowledge graph spanning sensors, P&IDs, ERP, and documents, with a multi-year transformation budget to back it up. Pick Pathnovo if your scope is engineering deliverables, instrument index, piping MTO, HAZOP register, mill cert traceability, EPC compliance, where the bottleneck is extracting structured data from PDF and scanned drawings within 48 hours at mid-EPC pricing. The three are not direct substitutes; many tier-1 owner-operators end up with two of the three, Predix or Cognite for operations plus Pathnovo upstream feeding either one.