Jacket leg · corrosion
LIVE
Asked
Which of last quarter’s inspections show corrosion, damage or encroachment?
Graphon
Reading1.1 Minspection photos40,000drone passes11 MentitiesEleven assets. Corrosion shows on the lower bushing housing in the September walkdown1, and the drone pass finds vegetation inside the four metre line2. The log records the corrosion spreading in 2024 with nothing signed since3, which leaves two assets overdue4

1.1 million inspection photos. Nobody has looked at most of them.

Graphon automatically builds and maintains
a large-scale ontology.

(graphons: a mathematical framework for analyzing large graphs, providing us with better recall and precision)

Graphon reads video, images, audio and documents, and goes into the content itself rather than the labels wrapped around it. It builds the ontology as it reads, roughly ten times deeper than anyone would model by hand.

So you are not held to broad entities like “a bottling line.” You can ask for amber 750 ml bottles with a tamper band, a lot code printed low on the shoulder and no allergen panel, and see where each result came from.

It rests on graphons, a mathematical framework for very large graphs. The founders developed the underlying work during their PhDs at Penn’s GRASP Lab.

You can drag the slider to see how it compares.

VideoImagesDocumentsTextStructuredPersonOrganizationLocationEventEquipmentPartEngine assemblyFitting the bracketSupplier of recordIdler bracketCowl fastener x38Cast mark HS-4820Part 11-4820-CTorque 48 NmMounting faceHeat lot 2247Mill certShipped 14 MarTier 3 · Halden SteelNot in any BOMDocumentsTextStructuredPersonOrganizationLocationEventontology stops herenot deep enough for agentic questions
Your graph doesn’t go deep.

It misses all the nuance, and no matter what you do it never quite works. More data only adds surface to guess from, and another ontology to model by hand.

Without Graphon3 sources · 4 entity types · stops at the entity
With Graphon5 sources · 6 entity types · 10 to 100 levels below

Ask both the same thing

One of them has read the footage. The other has read the filenames.

Two assistants, one question

Both were pointed at the same inspection programme. One indexed the reports. The other read the photography and the drone passes as well, and joined them to the asset register. The difference is not fluency. It is whether the answer can point at anything.

RAG · GraphRAG · vector search

Reads the text it managed to parse. Nothing to cite, because nothing was indexed.

Which of last quarter’s inspections show corrosion, damage or encroachment?
I can't answer that. The inspection report mentions corrosion findings, but the photography and drone passes aren't indexed, so I can't tell you the condition of any specific asset.
Graphon

Goes deep across video, images and documents, and returns a citation for every claim.

Which of last quarter’s inspections show corrosion, damage or encroachment?
Eleven assets. Corrosion shows on the lower bushing housing in the September walkdown1, and the drone pass finds vegetation inside the four metre line2. The log records the corrosion spreading in 2024 with nothing signed since3, which leaves two assets overdue4

Bring the question that keeps failing

Tell us what you have and the question your search keeps failing. Nothing to upload yet.

If it is a fit we will come back and say exactly what to send, then answer your question live with every claim cited.

  • Indexed in hours, not months
  • Every answer cited
  • Video, images and documents

Who usually joins

  • Director and VP of Operations, COO
  • Director of Asset Integrity, VP Engineering
  • Director and VP of Data and Analytics, Head of AI
  • Chief Digital Officer, VP Digital Transformation

Roughly what it costs you today. A range is fine.

SOC 2 Type II and ISO 27001 certifiedAll five trust services criteria · Visit the Trust Center