Data readiness

The 8th-largest theme in the research: 66 of 268 documents, 36 organizations. Decomposed into 5 topics, each with its own issues, evidence and steps.

66documents in the theme
12issues named
22sourced citations
5organizations cited
0sourced statistics

The topics

Each opens a dossier with the same shape: what the state of it is, the issues ranked by how many independent organizations name them - or marked as our own analysis where none does - the disagreements, the numbers, and the concrete steps under each remedy.

Unstructured data & retrieval

Most of what an enterprise knows is not in a table, and getting at it is its own engineering problem.

  • A document is treated as one object, and its meaning is lost at the first step 1
  • The unstructured problem is solved and the answer still cannot be trusted 1

77 documents · 37 organizations · 2 issues · 5 steps

Lineage, provenance & the golden source

Where a number came from, and whether two systems agree on it.

  • In a chain of agents, a weak signal upstream does not average out 2
  • Lineage is recorded at the door and lost through the pipeline ours
  • Two retrieval paths give two answers and neither is wrong ours

35 documents · 23 organizations · 3 issues · 8 steps

What "AI-ready" actually means

The distance between the data an organization holds and data a model can use.

  • Two thirds of the data is not in a state a model can use 2
  • Up to two fifths of the working week already goes on fixing the data 1
  • Remediating the estate, or remediating the paths the work runs on ours

28 documents · 18 organizations · 3 issues · 8 steps

Ownership & data products

Who is accountable for a dataset, and whether it is run as a product or a side effect.

  • The dataset everything depends on is owned by nobody 1
  • Each team rebuilds the same pipeline, and each one differs ours

17 documents · 10 organizations · 2 issues · 6 steps

Rights, consent & permitted use

Whether you are allowed to use the data this way, and where it may sit.

  • Access to a dataset is not the same as permission to train on it ours
  • Whether a deletion request reaches what has already been embedded ours

15 documents · 12 organizations · 2 issues · 6 steps

Where the sources disagree

No contradictions recorded yet.