VIGILIST 000001
Opposing counsel sends ten thousand documents. Overnight, Vigilist reads every one — coding each document and mapping every person, organization, and event into a knowledge graph. By morning your team has a review list that starts at the documents that decide the case, every call cited to the page that proves it.
VIGILIST 000002
Traditional e-discovery is priced for hundred-lawyer firms and still burns most of the budget on one thing: humans reading documents in order. Vigilist inverts that. The AI makes the first pass over everything, the knowledge graph shows you who and what matters, and your reviewers spend their hours on the documents that decide the case — inside one platform that also redacts, produces, and proves the work.
VIGILIST 000003
Relativity-style DAT/OPT productions with AI-assisted field mapping, PST/EML containers with families and threads intact, folders of PDFs and natives. Matters take multiple loads — their productions and your collection, side by side and labeled.
An LLM reads every document against your case criteria and codes it — relevant, key, or not — with confidence, reasoning, and excerpts. Attorneys stay in command with one-click overrides.
Blind control sets, recall and precision with confidence intervals, elusion testing over the discard pile. The statistics that turn "we used AI" into a number a judge accepts.
Full-text with phrases, booleans, and wildcards; semantic search when you don't know the words; hit reports with unique-hit counts for negotiating term lists with evidence.
Clustering, theme labels, and an AI case brief with citations that jump straight to the Bates number — your first read of a new production, done before you sit down.
Reason-coded redactions burned into every produced pixel. Second-pass QC approvals auto-invalidate if a redaction changes afterward — stale sign-offs structurally cannot ship.
Privilege tags derive withhold and redact-in-part dispositions; the log builds itself from safe metadata only — never from privileged content. Ready to serve as CSV.
Lock a set for gap-free Bates with families together; render endorsed PDF or single-page TIFF; carve out natives; split volumes by size; deliver Concordance DAT/OPT any platform can load.
Locking runs validation first: unapproved redactions, privileged documents going out clean, received documents headed back to their producer. Resolve — or override on the record.
Every document tells its story: source, hashes, extraction, tags, decisions, redactions, productions. Matter-level custody reports and an honest exceptions ledger, one click away.
Keyboard-driven coding, bulk tagging across thousands of documents, queues and batches with statistical QC, and an in-browser viewer for images, text, natives, and media.
SSO with your firm's identity provider (SAML/OIDC), enforceable per organization; role-based access; an append-only audit trail behind every action.
VIGILIST 000004
Vigilist reads the corpus and builds the cast of characters: every person and organization, resolved across aliases — "J. Rivera," "Rivera, Jorge," and the email address are one person, with a review queue for close calls so a human confirms every merge.
Click any name, anywhere: an AI-written profile, every mention in context, and who they're connected to — employer, counsel, correspondent, party to an agreement — each relationship backed by the phrase in the document that proves it. Key players rank by presence, so the people who matter surface on day one.
Events — meetings, payments, agreements — are extracted from day one, building toward matter timelines and relationship graphs on the same foundation.
VIGILIST 000005
Drop in their production or your client's mailboxes. Metadata, families, threads, hashes, and OCR handled automatically — keep adding loads as the case grows.
The AI codes the corpus and briefs the team; reviewers work the queues at keyboard speed. Validate the machine with control sets whenever the numbers need to hold.
Draw redactions, QC them, and let the privilege log build itself from your privilege calls — dispositions, descriptions, and all.
Lock the set, stamp Bates, render, package DAT/OPT with a hashed manifest. Hit reports, sampling statistics, and custody reports stand behind every choice.
VIGILIST 000006
Every claim comes with its evidence attached. Redaction approvals that expire when the redaction changes. Production locks that record who overrode which warning. Frozen random samples with Wilson confidence intervals. Validation reports that persist exactly as run. An append-only audit log under all of it.
"How do you know the AI didn't miss anything?"
RECALL 87.4% [CI 81.2–92.1] · ELUSION 0.4% · EST. MISSED 38–61 OF 9,912 · REPORT #12, PRESERVED AS RUN
— the answer, from a sample Vigilist validation report
VIGILIST 000007
- "For twenty years I watched the side with the
- bigger discovery budget start every case ahead —
- before the merits ever got a hearing. Vigilist is
- the first platform that ends that head start. My
- team ingests a production on Friday and argues from
- it on Monday — and when the other side questions
- the process, the numbers are already in the record."
VIGILIST 000008
Billed on the gigabytes you upload — not the "processed" gigabytes other platforms expand them into. AI review, entity intelligence, productions, and defensibility included. No seats, no minimums, no lock-in.
$90/month · roughly 25,000 documents — read, coded, and mapped
| Platform | To get a price | A 5 GB matter | AI review | Commitments |
|---|---|---|---|---|
| Vigilist | You just did — $18 / GB / mo, published | $90 / month, everything included. Start today. | Every document read and coded overnight, with validation statistics built for court | None — month to month, billed on the raw GB you upload |
| RelativityOne | Sales quote, through a service provider | Processing, seat & hosting fees — est. $300–700 in month one, before anyone reads a page | aiR — in whatever package your provider negotiates | Annual terms typical · billed on processed data and peak review volume |
| DISCO | Sales quote — pricing not published | Accounts reported at $20k+ / yr; a small matter may not clear the door | Cecilia — in whatever bundle you negotiate | Billed on processed data — more GB than you sent |
| Everlaw | Sales quote — pricing not published | Est. $2,000–5,000 / mo platform fee before the first gigabyte | Batch AI actions sold separately, as credits | The platform fee is the minimum |
Competitor figures are third-party estimates of quote-only pricing and Winter 2026 industry-survey data, as of July 2026. They don't publish prices. We do.
Vigilist reads the corpus into a cast of characters — every person, organization, and event, resolved across aliases, every relationship cited to the document that proves it. Click a name; see the case.
Every call ships with confidence, reasoning, and quoted excerpts — validated by control sets, recall, precision, and elusion statistics preserved exactly as run. When they ask how you know it didn't miss anything, you hand over a number.
Not a chat assistant you consult one document at a time, and not a batch action metered in credits. The first pass over everything is the product — it runs the night your data lands.
VIGILIST 000009
Bring a sample production and we'll walk the whole lifecycle — ingest, AI review, redaction, and a Bates-stamped production back out. Start to finish.
VIGILIST · DEMO REQUEST
Received.
We'll reply within one business day to set up a walkthrough on your own production.