Search the whole room
Hybrid meaning + keyword search across every document, with grounded answers that cite their sources — surface the terms, obligations, and exceptions that matter without reading linearly.
Due Diligence · Buy-Side & Third-Party Risk
Read the data room, then vet who's behind it.
Private, on-premise diligence: search a document set you've been given, map the entities and relationships inside it, then extend selected names into authorized public-source research — all on hardware you control. Confidential deal and counterparty material never touches the cloud.
Private diligence on your own computer. Search a set of documents you've been handed, map the people and companies inside it and how they connect, then check selected names against public records — all without the confidential deal material ever going to the cloud.
On-premise
No cloud upload
Entity mapping
Authorized OSINT
Cited answers
The Problem
Two jobs, and neither fits a public tool.
Diligence is read everything in the room to find the red flags — and understand who you're actually dealing with. The documents arrive under NDA, so uploading them to a cloud AI tool is often off the table; and a document tool alone can't tell you what the outside world knows about the parties.
Diligence means reading everything in the room to spot the red flags — and figuring out who you're really dealing with. The documents come under NDA, so uploading them to a cloud AI tool is usually off the table; and a document tool alone can't tell you what the outside world knows about the parties.
Confidential in; nothing leaks out.Deal and counterparty documents are exactly the material you can't paste into a public model that may retain it. The analysis has to happen inside your own environment.
Confidential in; nothing leaks out.Deal and counterparty documents are exactly what you can't paste into a public tool that might keep it. The work has to happen inside your own environment.
The Difference
Inside the documents — and beyond them.
BDS handles the received set; the OSINT extension handles the parties. Same private environment, one workflow.
BDS handles the documents you were given; the OSINT extension checks out the parties. Same private setup, one workflow.
Map the entities
The people, companies, and relationships inside the set, as a knowledge graph — so you see the structure and the connections, not just a file list.
Vet the parties
Extend selected entities into authorized public-source OSINT — turning a name from the documents into structured, triaged external evidence for a reviewer to judge.
Auditable evidence trails
Every answer and lead traces back to its source, so a diligence memo rests on inspectable evidence, not a black box.
Search the whole room
Search every document by meaning and by keyword, with answers that quote the source — surface the key terms, obligations, and exceptions without reading page by page.
Map the players
See the people, companies, and relationships inside the set as a map — the structure and connections, not just a file list.
Check out the parties
Take a selected name and check it against authorized public records (OSINT) — turning a name in the documents into organized outside evidence for a reviewer to weigh.
Every lead is traceable
Every answer and lead points back to its source, so a diligence memo rests on evidence you can inspect — not a black box.
OSINT is the one outbound step — and you control it.Reading and mapping your documents is fully local. The OSINT extension is a separate step you invoke deliberately; when you use it, only the entity identifiers you choose (a name, company, or domain) are looked up against public sources — never your confidential documents. Route those lookups through your own egress allowlist so every request is controlled and logged, or leave OSINT disabled for a fully air-gapped run.
OSINT is the one step that reaches out — and you control it.Reading and mapping your documents stays entirely on your computer. The OSINT check is a separate step you choose to run; when you do, only the names you pick (a person, company, or website) are looked up against public sources — never your confidential documents. You can route those lookups through your own controlled connection and log them, or leave OSINT off for a fully offline run.
BDS is the private intelligence layer over a document set you've received — it makes the room searchable, mapped, and answerable on your own hardware. It's not a shared virtual data room you upload to; it runs in your environment, for your side of the deal.
BDS sits over a set of documents you've been handed — making the room searchable, mapped, and answerable on your own computer. It's not a shared online data room you upload to; it runs in your environment, for your side of the deal.
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 is closed to the network by default and makes no outbound calls — the one exception is the OSINT module, which you enable deliberately and route through your own egress. See the deployment & security model →
BDS runs entirely on your own hardware. All the AI and storage are local, and the core system makes no internet connections at all — the one exception is the OSINT check, which you turn on deliberately and route through your own connection. See how it's deployed and secured →
Proof
The same pipeline, shown on real corpora.
The hybrid search, grounded answers, and entity graph shown on real litigation and corporate-fraud document sets are the same capabilities a diligence set needs — a body of documents, read, mapped, and questioned. Watch it run and judge for yourself.
The search, cited answers, and map shown on real litigation and corporate-fraud documents are exactly what a diligence set needs — a pile of documents, read, mapped, and questioned. Watch it run and judge for yourself.
A Living Archive
And it's cumulative. Keep adding documents as the deal or subject develops and they join the existing corpus — so a file that grows over months and years stays one searchable, permanent archive.
And it keeps growing. Add documents over time and they join what's already there — so the whole file stays one searchable, permanent archive.
A Real Case, Not a Demo · US v. Google (ad-tech)
We ran BDS on the US v. Google ad-tech docket.
The US v. Google ad-tech antitrust docket (DOJ) is a large public record of trial exhibits — no synthetic data. Here's what one real run produced.
280
documents
6,773
companies, people & roles
53,800
connections found
292
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.
Which Google products are central to the case?
Its ad exchange (AdX), its demand-side platform (DoubleClick Bid Manager), and DoubleClick — the tools the DOJ says Google tied together.
Source: DOJ trial exhibits
What is the DOJ's core allegation?
That Google controls both the buy side and the sell side of digital advertising by tying its demand-side platform to its ad exchange.
Source: DOJ trial exhibits
Which competitors are named?
Amazon, The Trade Desk, AppNexus, Rubicon and PubMatic, among others.
Source: DOJ trial exhibits
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 6,773 entities and 53,800 connections — far too dense to read on a page. So this shows the strongest 16-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 set.
Send a representative, de-identified sample — one deal or subject, 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.
Send us a sample of your own documents (with sensitive details removed) — one deal or subject, 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, built from your material. A full setup runs entirely on your hardware.
Other practices: Private eDiscovery · Investigations · Real Estate