Compare the storage your deployment actually needs
Access arranged as part of a pilot
Analect Store is the storage analysis workflow in the Analect platform. It measures every form your material could be kept in, in actual bytes, raw and compressed, and models both scenarios: keeping your source documents alongside the fragment layer, and replacing them. It distinguishes bytes reduced from bill reduced, and returns a verdict that is allowed to be no.
The storage bill nobody budgeted
Most stored enterprise data is unstructured, and AI has so far made that worse rather than better: a retrieval pipeline keeps the original, then a chunk store, then embeddings, then an index. The standard answer to the unstructured data problem is an architecture that makes the estate bigger. Store exists to weigh that honestly, in bytes, including the forms nobody budgeted for.
Weigh, plan, verdict
Weigh the forms
Every way the material could be kept, in bytes, raw and compressed: the source, the fragments, and everything in between.
Model both scenarios
Corpus size, copies, compression, and the question that decides everything: do the source documents stay? Store models retain-source and replace-source separately, because they give opposite answers.
Read the verdict
A storage bill per tier, the assumptions behind the compression figures, and an explicit split between bytes reduced and bill reduced. The verdict is allowed to be no.
What we measured
What that figure does and does not cover
- It is the fragment layer measured against one source file: roughly 300 KB against 3.5 MB.
- It excludes source retention, metadata, indexes, replicas and backups. Include those and the comparison changes.
- Keeping your sources — which you should, because fragments cite them — means total retained storage goes up, not down. That is the retain-source scenario, and Store prices it.
- Bytes reduced is not the same as bill reduced. Storage pricing is per tier, and a smaller object in a more expensive tier can cost more.
Weighing happens in the browser. Nothing leaves your machine on this screen.
What the storage verdict contains
Two scenarios, priced separately, because they give opposite answers and conflating them is how storage claims become misleading.
| Field | What it contains |
|---|---|
| Source documents | Raw and compressed. In the retain-source scenario these stay, and they dominate the total. |
| Fragment layer | The structured tables. This is the figure that is small; it is not the whole picture. |
| Indexes and metadata | What is needed to retrieve a fragment, which is a real cost and is frequently left out of comparisons. |
| Replicas and backups | Multiplied by your actual retention policy, not by a best case. |
| Tier and price | Per storage tier, because a smaller object in a more expensive tier can cost more. |
| Bytes against bill | Reported as two separate numbers. A reduction in bytes is not automatically a reduction in bill. |
This is the shape of the report, not a result. The figures in yours come from your own documents.
What you can change, and what moves when you do
Every figure rests on these. They are editable rather than baked in, because a model whose assumptions you cannot see or change is not evidence.
| Assumption | Effect |
|---|---|
| Do the sources stay? | The question that decides the verdict. Keep them — and you should, because fragments cite them — and total storage goes up, not down. |
| Compression ratio | Stated per format rather than assumed. Already-compressed formats have little left to give. |
| Copies and retention | Replicas and backup generations multiply every line above. |
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.
- Your retention policy, and how many copies of each thing it actually implies
- Which storage tiers you are on and what each costs you
- Whether the sources are being retained, archived to a colder tier, or kept as they are
Weigh your estate.
A pilot measures every form your material could be kept in, models the retention plan and returns a storage verdict, including the answer that says no.