Organize
Batch document sets are sorted into usable research collections before deeper processing begins.
BDS Platform
Batch Document Search
A local, private document-intelligence system that organizes complex files, indexes them for hybrid search, extracts entities, and builds an Obsidian-compatible knowledge graph.
A private tool that reads a big pile of your documents, lets you search them by meaning — not just matching keywords — and draws a map of the people and organizations inside them. It runs on your own computer, so nothing is uploaded anywhere.
The Pipeline
Four phases, from raw files to a knowledge graph.
Four steps, from a pile of files to a searchable map.
BDS moves a document set through four stages — and each one is usable before the next completes.
Your documents move through four steps — and each step is useful before the next one finishes.
Index
After Phase 2, hybrid BM25 and semantic search become available for immediate review.
Extract Entities
Phase 3a identifies people, organizations, places, and key terms for structured research review.
Build Knowledge Graphs
Phase 3b expands entity cards, relationships, and Obsidian-compatible graph exploration.
1 · Sort
Your files are sorted into tidy collections so everything related sits together.
2 · Make searchable
Once sorted, you can search them right away — by meaning and by exact words.
3 · Find the players
It picks out the people, organizations, and places named across the documents.
4 · Connect them
It draws a clickable map of how those people and organizations relate.
A Living Archive
Cumulative by design.
It grows with the matter.
Add documents to a matter as it develops and they join the existing corpus — earlier material stays indexed and searchable, and isn't reprocessed. A matter becomes a growing, permanent archive rather than a one-time load.
As a matter grows, drop in new documents and they join what's already there — the older ones stay searchable and don't have to be redone. Your case file becomes one living, permanent archive, not a one-time import.
How To Read Results
Search signals, not magic scores.
A quick guide to the relevance signals BDS shows beside every result.
Next to every result, BDS tells you why it showed up — so you know how much to trust it.
"Both" — the strongest signThe wording matched and the meaning matched. Trust these first.
Meaning-only, or words-onlyOnly the meaning lined up, or only the exact words did. Useful — just read with a little more care.
"Probably not here"BDS warns you when your documents likely don't hold the answer — instead of making one up.
| Signal | Plain-English meaning |
|---|---|
| both | Meaning-match and word-match agree. This is the strongest relevance signal. |
| vector | Meaning-match. The text reads like the question even when different words are used. |
| bm25 | Word-match. It shares query words but may not actually answer the question. |
| CORPUS_GAP | The documents may not contain strong evidence for the question. |
Watch It Run · Ingestion
A real case, ingested end to end.
A 30-document random sample from the SEC v. Ripple docket moves through the pipeline — organize and OCR, chunk and embed, then entity extraction — captured locally at real speed. The same run produces the knowledge graph below.
Watch a real set of 30 court documents get read, sorted, and understood by BDS — start to finish, on one computer, at real speed. The same run builds the map shown further down.
Local capture: a 30-document SEC v. Ripple sample ingested end to end — organize/OCR, indexing, and entity extraction.
Hybrid Search · Synthesis
Ask the corpus. Read the signals.
Once a sample is indexed it's searchable. This capture runs hybrid semantic + keyword queries across two projects — the SEC v. Ripple and Enron corpora — showing each result row with the relevance signal behind it, then synthesizes an answer drawn only from the retrieved passages.
Here BDS answers questions about the documents — searching by meaning and by keyword — and writes the answer using only what's actually in the files, with the sources shown. If the answer isn't there, it says so.
Local capture: hybrid semantic + keyword search across the SEC v. Ripple and Enron projects, result rows with relevance signals, and grounded answer synthesis.
Phase 3b · Knowledge Graph
The extracted entities, connected.
Once entities are pulled from the corpus, BDS writes them out as an Obsidian-compatible vault. This capture walks the same 30-document SEC v. Ripple sample: each node is a person, organization, or place, and each line is a relationship found across the filings — the same evidence, now navigable.
This is the map: every dot is a person, organization, or place found in the documents, and every line joins two that appear together. You can pan around it and click any one to see what it's connected to.
Local capture: the SEC v. Ripple entity graph in Obsidian — panning the full graph, then isolating a single entity to follow its links.
Try It On Your Documents
See BDS run on a sample of your own files.
Send a representative set — one subject, roughly 100–300 documents, any common format, scanned PDFs included — and a BDS evaluation returns searchable results, an entity graph, and entity profiles drawn from your own material.
Send us a sample of your own documents — one topic, around 100–300 files, scans are fine — and we'll show you the search, the map of who's involved, and short profiles of the key players, all built from your material.
For your work: Private eDiscovery · Investigations · Due Diligence · Real Estate