An entity knowledge graph
People, organizations, and other entities pulled from the whole set, with how they connect and co-occur across emails, filings, and memos — the network view a document-by-document read can't give you.
Investigations · Compliance & Regulatory
See who connects to whom — without the file leaving the building.
Private, on-premise AI for confidential investigations: search the whole document set, map the people and organizations and how they relate, and get answers that cite their sources — on hardware you control. Internal, regulatory, and fraud matters never touch a third-party model.
Private AI for confidential investigations, running on your own computer. Search the whole set of documents, see a map of the people and organizations and how they're linked, and get answers that quote the source. Internal, regulatory, and fraud matters never touch an outside service.
On-premise
No cloud upload
Entity relationship graph
Cited answers
Auditable evidence trails
The Problem
Investigation material is the last thing you can afford to leak.
An internal or regulatory investigation is defined by its sensitivity — subjects, whistleblowers, privileged advice, unproven allegations. Pasting any of it into a public AI tool routes it through a third party that may retain the input, and can compromise confidentiality before the matter is even understood.
An investigation is defined by how sensitive it is — subjects, whistleblowers, privileged advice, allegations that aren't proven. Paste any of it into a public AI tool and it passes through an outside company that may keep it — a confidentiality breach you caused yourself, before you even understand the matter.
Public AI is an exposure, not a tool.Consumer platforms often keep prompts and documents to train their models. For an investigation, that's a disclosure you can't take back — and a confidentiality problem you created yourself.
Public AI is a leak, not a tool.Consumer tools often keep what you give them to train themselves. In an investigation, that's a disclosure you can't undo — and a confidentiality problem you created.
The safe path is the obvious one: run the AI on the investigation, not on someone else's servers.
The Difference
The relationships a linear review misses.
An investigation is rarely a keyword hunt — it's a question of who knew what, who spoke to whom, and when. That's exactly what BDS surfaces.
An investigation is rarely a keyword hunt — it's about who knew what, who talked to whom, and when. That's exactly what BDS surfaces.
Grounded answers that cite sources
Ask the record a question and get an answer built only from retrieved passages, with citations — and a plain "not supported" when the documents don't back it. Where to look, not what to conclude.
Hybrid search over everything
Meaning-match and word-match run together across the full corpus, with a signal on each result so you know how much to trust it before you open it.
Auditable evidence trails
Every lead traces back to its source document, so a reviewer can inspect how it was produced — defensible, not a black box.
A map of who's connected
It pulls the people and organizations out of the whole set and shows how they link up across emails, filings, and memos — the network view you can't get reading one document at a time.
Answers that quote the source
Ask the record a question and get an answer built only from the documents, with the passages — and a plain "not supported" when they don't back it.
Search across everything at once
It searches by meaning and by exact words across the whole set, and tells you how strong each match is before you open it.
Every lead is traceable
Every lead points back to its source document, so a reviewer can see how it was found — defensible, not a black box.
The Private Answer
Run entirely on your own hardware.
BDS is on-premise software. All AI inference and all storage are local; the core pipeline makes no outbound network calls and is closed to the network by default — it can run fully air-gapped. Nothing about a matter transits a third party. See the deployment & security model →
BDS runs entirely on your own hardware. All the AI and all the storage are local, the core system makes no internet connections, and it can run with no internet at all. Nothing about a matter passes through anyone else. See how it's deployed and secured →
Proof
Demonstrated on the canonical corporate-fraud corpus.
We ran BDS on the Enron corpus — the reference set for investigations, all emails and filings — building the entity graph and running hybrid search with grounded answers, captured locally at real speed. The same pipeline runs on your matter, on your hardware.
We ran BDS on the Enron files — the classic reference set for investigations, all emails and filings — building the map of who's involved and searching it with cited answers, recorded live on one computer. The same thing runs on your matter, on your hardware.
A Living Archive
And it's cumulative. Keep adding documents as the matter develops and they join the existing corpus — so a file that grows over years stays one searchable, permanent archive.
And it keeps growing. Add documents over the years and they join what's already there — so the whole file stays one searchable, permanent archive.
A Real Case, Not a Demo · Enron
We ran BDS on the Enron corpus.
The Enron corpus is the reference set for corporate-fraud investigations — real emails and filings. Here's what one real run produced.
29
documents
9,151
people, companies & entities
58,823
connections found
470
entity profiles
One real matter, shown to illustrate the pipeline — an illustrative case, not a benchmark average.
Real questions, answered from the documents.
Three questions put to the case file and answered by BDS from the source documents — each answer built only from what the record says, with the source shown. Nothing here is hand-written.
What were the LJM partnerships?
Off-the-books entities (LJM Cayman and LJM2) created by Enron in 1999 and run by CFO Andrew Fastow — used to hide debt and inflate results.
Source: Bankruptcy examiner filings
Which Enron entities appear across the files?
Enron and its subsidiaries — Enron North America, Enron Power Marketing, Enron Energy Services, Transwestern and Portland General Electric, among others.
Source: Bankruptcy filings
What was FERC's role?
The Federal Energy Regulatory Commission oversaw Enron's regulated energy operations — pipelines and utilities — setting rates and reviewing compliance.
Source: Bankruptcy filings
Grounded answers from hybrid search — where to look in the record, not a verdict on it.
The Map
The people, companies, and connections — mapped.
A readable slice of the entity map from the same run — the Obsidian-style vault BDS builds: each dot is an entity found in the documents, sized by how often it appears; each line joins two that show up together. Tap any dot to drill into it, or use a question above to light up its cast; drag to pan, scroll to zoom.
You’re seeing a readable slice — not the whole graph.The full map from this run holds 9,151 entities and 58,823 connections — far too dense to read on a page. So this shows the strongest 14-entity backbone: the people, organizations and places that come up most, and the ties between them. Drag, zoom, and click to explore — the full graph lives inside the system, on your hardware.
Every dot and line is drawn from the source documents — nothing is invented.
Private AI. On Your Hardware.
Try it on a sample of your own matter.
Send a representative, de-identified sample — one matter, roughly 100–300 documents, scanned PDFs included — and a BDS evaluation returns searchable results, the entity graph, and entity profiles drawn from your own material. A real deployment runs entirely on-premise; your live matter stays in your environment.
Send us a sample of your own documents (with sensitive details removed) — one matter, 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 people, built from your material. A full setup runs entirely on your hardware; your live matter stays with you.
Other practices: Private eDiscovery · Due Diligence · Real Estate