Inspect the knowledge extracted from your documents
Access arranged as part of a pilot
Analect Explorer is the decomposition and exploration workflow in the Analect platform. It reads unstructured documents and produces addressable knowledge fragments, then lets you explore what the material actually says: entities, relationships, connections between any two things, and the gaps.
From a file to a map of what it says
Upload
Drop a file or paste a public URL. Large documents are handled in sections automatically.
Understand
Headline structure, entities, key statements and the shape of the document at a glance.
Connect
The knowledge graph: paths, hubs and bridges between any two concepts.
Project
Readings across the loaded material: what shifts, what repeats, what is missing.
Visualise
Boards and visuals built from what the analysis found.
Analyses that outlive the session
Save analyzes to a library and come back to them from any session. Question the fragments through a chosen lens and get answers with citations to the exact source sentences, so a fragment that has been extracted wrongly can be traced back and corrected rather than quietly believed. Hand a loaded analysis straight to Forecast, Proof or Store without re-uploading anything.
Your documents stay where they are
Decomposition runs inside your environment, and only the fragments a question needs are sent to whichever model you have chosen, under that provider's terms. That reduces how much of your text leaves your estate. It is not by itself anonymisation or a security boundary, and where each processing step happens is confirmed for your deployment rather than assumed.
What decomposition produces
A sentence in a document becomes a small number of structured statements, each carrying a pointer back to where it came from.
| Field | What it contains |
|---|---|
| The assertion | What the document states, as a structured statement rather than a span of prose. |
| Entities and relationships | What the statement is about, and how those things relate — the part that makes a question answerable without reading everything. |
| Source pointer | Document, location and the sentence itself. Every fragment carries one; a fragment that cannot cite its source is a defect, not a result. |
| Confidence | How certain the extraction is, so low-confidence material can be reviewed rather than silently trusted. |
This is the shape of the report, not a result. The figures in yours come from your own documents.
What we will not claim in advance
These are answered by running this on your own documents, not by a specification sheet. They are the things most likely to decide whether the approach works for you, which is exactly why guessing at them would be unhelpful.
- Which of your document formats decompose cleanly, and which need pre-processing. This is the single biggest determinant of whether the approach works on your estate, so we measure it rather than claim it
- How your tables, figures and scanned material are handled — these behave very differently from prose and results do not transfer between them
- How updates are handled: which documents change, how often, and what re-processing costs
- The correction route when a fragment is wrong — who reviews it, and what happens to answers that already cited it
See your documents decomposed.
A pilot decomposes a slice of your estate so you can explore what the material actually says, with every fragment tied back to the sentence it came from.