AI due diligence · built for deal teams
Turn a data room into
a verified diligence workstream.
Evidence-backed diligence execution, not document chat. Vaultrix analyzes the entire transaction data room, maps evidence to your DD request list, flags gaps and conflicts, and links every finding to the exact source — giving deal teams a traceable first pass instead of starting from a blank spreadsheet.
Built for investment banking, private equity, M&A, credit, transaction services, and corporate development teams.
No card required for the free tier.
AI due diligence, defined
What is AI due diligence?
AI due diligence is the use of AI to read a transaction data room, answer diligence questions with citations back to the exact source, and flag what evidence is missing or conflicting — instead of a team manually reading every file. Vaultrix applies this directly to your deal: upload the room, map it against your request list, surface the evidence for each item, flag what's missing or conflicting, verify the citations behind every high-risk answer, and route the findings to your team for sign-off. The AI performs the first pass; nothing counts as closed until a person reviews it.
See how Vaultrix's AI due diligence software works →Explore diligence
workflows.
Deeper guides on how each part of the workflow actually works.
AI Due Diligence
What it means, how it works, and its limitations.
M&A Due Diligence
The process, the request list, and reporting findings.
Financial Due Diligence
Revenue, EBITDA, debt, and cross-document consistency.
Request Lists
Request vs. checklist, and the four evidence outcomes.
AI Data Rooms
How an AI layer differs from a traditional VDR.
Why teams use Vaultrix
Diligence is not a search problem. It is an evidence problem. Anyone can summarise a data room. Vaultrix determines what the documents actually support, and gets a human to sign off on it.
Upload the transaction data room once and reason across the complete document set.
Surface requests with no evidence, partial support, or conflicting information.
Every important finding traces back to the underlying source — a page, a slide, or a spreadsheet cell.
AI proposes. The deal team reviews and signs off before anything counts as closed.
Revenue, FY22 to FY26
Revenue grew from ₹700 Cr to ₹850 Cr year over year, with EBITDA margin expanding from 17.0% to 18.2% over the same period.
From data room to a verified diligence workstream.
Read the whole room.
Reason across the entire data room, not one document at a time.
Trace every finding.
Findings trace to exact pages, slides, sheets, and cells.
Surface what's missing.
Missing, partially supported, and conflicting evidence gets flagged.
Humans decide.
A reviewer approves each finding before it's treated as closed, with threaded comments and a recorded reason for every rejection.
Close the loop.
Everything maps back to the DD request list, then out to a report deck, workbook, or One Pager.
From upload to
fully indexed.
Vaultrix parses, indexes, and categorises every file in the room, so nothing has to be read in manually before analysis can start.
PDF, Word, Excel, PowerPoint, CSV, and text files, uploaded once as a single ZIP.
Diligence request list
Your DD request list, pre-answered.
Upload the diligence request list you already use. Vaultrix maps each request to the evidence in the data room and proposes a status: evidence found, partially supported, or no evidence found.
Don't just find what's in the room. Find what's missing. Every request gets one of four outcomes: evidence found, partially supported, no evidence found, or conflicting, so gaps surface during diligence instead of after closing.
Turn every gap into a management question list, one question per open item and numbered by workstream, and export the tracker to Excel when you need to share where things stand.
See it in action
Watch the analyst work through a live deal.
Answer → Evidence → Verification. Click a citation below to see the connection.
Cross-referencing the audited financial statements and transaction model in the data room:
- FY25: ₹119.0 Cr
- FY26: ₹155.0 Cr
- Year-over-year expansion: +30.3%
- FY25: 17.0% · FY26: 18.2% (+120 bps)
The audit indicates the expansion was driven by operating leverage on higher factory utilization and a mix shift toward higher-margin assemblies. Revenue grew from ₹700 Cr to ₹850 Cr .
Click a citation badge to inspect the exact cell or page it came from →
| ₹ Cr | FY22 | FY23 | FY24 | FY25 | FY26 |
|---|---|---|---|---|---|
| Revenue | 420 | 510 | 605 | 700 | 850 |
| EBITDA | 63 | 79 | 96 | 119 | 155 |
| Margin | 15.0% | 15.5% | 15.9% | 17.0% | 18.2% |
Direct extraction from the financial model's cell matrix.
| # | Diligence request | Category | AI finding | Banker sign-off |
|---|---|---|---|---|
| 01 | Transfer pricing study, if applicable to intercompany transactions | Tax / Finance | Evidence found (high) Matches Transfer_Pricing_Study.pdf (p.21) |
|
| 02 | Audited standalone and consolidated financial statements, FY24 to FY26 | Financial | Evidence found (high) Found 3 statutory audit reports (100% complete) |
|
| 03 | Schedule of all outstanding borrowings, terms, covenants, and collateral | Contracts | Partially supported Apex & Continental agreements located; one pledge missing |
|
| 04 | Details of all material litigation, arbitration, or regulatory proceedings | Legal | Partially supported $1.8M breach-of-contract matter identified; docket pending |
|
| 05 | Confirmation of compliance with applicable regulatory filing requirements | Corporate | No evidence found No filings uploaded to the Corporate folder |
Governance rule: the AI proposes findings with citations. An item only closes once a verified banker confirms it.
| Document | Category | Pages | Status | Action |
|---|---|---|---|---|
| XLSXFive_Year_Financial_Model.xlsx | Financial | 3 sheets | Parsed | |
| PDFStatutory_Audit_Report_FY26.pdf | Financial | 22 pages | Parsed | |
| PDFCredit_Agreement_Apex_Finance.pdf | Contracts | 24 pages | Parsed | |
| PDFCredit_Agreement_Continental.pdf | Contracts | 21 pages | Parsed | |
| DOCXMemorandum_of_Association.docx | Corporate | 28 pages | Parsed | |
| PDFDraft_Red_Herring_Prospectus.pdf | Regulatory | 180 pages | Parsed | |
| PDFTransfer_Pricing_Study.pdf | Tax | 42 pages | Parsed |
The Company is currently a defendant in one litigation matter alleging breach of contract, with claim amount of $1.8M…
No default or material dispute exists under any major supply agreement except the disclosed proceeding alleging breach of contract…
Consolidated EBITDA margin projected at 18.2% for FY26 vs. 17.0% in FY25, an increase of 120 bps…
The Borrower certifies that, except as set forth on Schedule 8.4, there are no actions, suits, or proceedings pending or threatened against the Borrower…
Ask the questions that
normally take hours.
"Summarise working capital movements over the last three fiscal years."
"What percentage of revenue comes from the top 10 customers?"
"List all outstanding borrowings, maturity dates, covenants, and collateral."
"Identify change of control provisions across material agreements."
"Find evidence of transfer pricing documentation and identify gaps."
"The financial model says X. Does the audited financial statement support it?"
"How do the target's EBITDA margins compare with listed peers, and has anything changed in the sector this year?"
Deliverables
From answers to
the files you actually send.
Draft the deliverables and working papers a deal team owes, straight from the data room, in your firm's branding. Every figure stays traceable to its source in Vaultrix, and a reviewer signs off before anything goes out.
FDD report deck
A financial due diligence report in PowerPoint: an Executive Report (about 10 to 12 slides) or a Full FDD Report (about 90 to 100 slides), with headline tables, commentary, and EBITDA, net debt, and working capital bridges. Full reports start from a section plan you approve. Download as .pptx or Word.
Excel workbooks
Ask for a schedule, comparison, or tracker and get a multi-sheet .xlsx with live cross-sheet formulas. Numbers pulled from source spreadsheets are re-checked against the original cells, and any mismatch is flagged instead of passed through.
One Pager
A one-page deal overview of the target, drafted from the room and downloadable as a PDF for an IC pack or a first internal read.
Red-flag memo
The first-pass note on where to spend the hours: the top risks ranked from request-list statuses, conflicts, and open management questions, each with its impact and next step. Built from what the room already shows, with no new AI findings, and marked preliminary on every page. Remove or restore individual risks, then export as PDF or Word.
QofE adjustments schedule
Proposed quality-of-earnings adjustments, each checked against the file, page, or cell it came from. Only the rows you accept count toward adjusted EBITDA, and a mismatch with management's own figure is raised as a conflict. Export to Excel, and the Full FDD Report's bridge uses the accepted rows.
Management question list
One question for every request with no evidence, partial evidence, or a conflict, numbered by workstream. Your edits survive regeneration. Export as Word or Excel, ready to send to management.
Request-list tracker
Export the request list to Excel (all items, open items, or conflicts only) with a live status summary and AI-proposed statuses clearly labelled.
Your branding, clean exports
Decks, Word and PDF files, the One Pager, workbooks, and the question list carry your firm's name, logo, colours, fonts, legal line, and disclaimer. Citations stay in the app, where a reviewer can click through to the exact page or cell, so exported files are clean and ready to send.
Security capabilities available today
Built for the most
confidential room in the building.
Vaultrix is built for teams handling unannounced transactions.
Data isolation
Retrieval, uploads, and document access are scoped to a single deal, enforced server-side. One deal can never read another's data.
Private processing
Document indexing and embedding happen on Vaultrix infrastructure — text never leaves our servers for that step.
Encryption
Data is encrypted at rest and in transit using standard cloud-provider encryption. No dedicated per-deal keys yet.
Access control
A user needs an explicit membership on a deal — not just membership in your organisation — before they can query it, upload to it, or view its documents.
Activity logging
Questions, document views, uploads, exports, and review decisions are logged with who and when. A deal admin can export a deal’s log as CSV or Excel, filtered by date range.
Deletion
Deleting a deal removes its files from storage and its rows from the database, permanently. No certificate, just verified deletion.
Built specifically for diligence workflows
Your data room stores the evidence. Vaultrix works through it.
A data room gives you storage. A general-purpose AI tool gives you a conversation. Neither is purpose-built for the full DD workflow — mapping a request list to evidence, flagging gaps and conflicts, routing findings for sign-off, and drafting the report deck, workbook, and One Pager at the end. Vaultrix complements your existing diligence stack rather than asking you to replace it.
| Capability | Vaultrix | Traditional data room | General-purpose AI |
|---|---|---|---|
| Whole data-room ingestion | ✓ Whole ZIP archive, one upload | ✓ Stores files, no reading them for you | ✗ Not purpose-built for the full DD workflow |
| Scanned PDFs | ✓ OCR on image-only pages, cited to the page like any native PDF | ~ Some offer OCR for search only | ~ Varies by tool, one file at a time |
| Cross-document reasoning | ✓ Reasons across the whole data room | ✗ No synthesis, you read every file yourself | ✗ Not purpose-built for the full DD workflow |
| Exact source citations | ✓ Every answer cites the page, slide, or cell | ✗ Stores and organizes evidence; does not perform the diligence analysis layer | ✗ Not purpose-built for the full DD workflow |
| DD request-list mapping | ✓ Auto-maps your DD request list to uploads | ✗ Manual spreadsheet tracking | ✗ Not purpose-built for the full DD workflow |
| Missing evidence detection | ✓ Flags found, partial, and no-evidence items | ✗ Manual review to spot gaps | ✗ Not purpose-built for the full DD workflow |
| Conflict detection | ✓ Flags contradicting information across documents | ✗ No detection capability | ✗ Not purpose-built for the full DD workflow |
| Live market context | ✓ Optional Research mode adds live web search, labeled apart from data-room evidence | ✗ Only what's been uploaded | ~ Web search, but not alongside your data room |
| FDD report deck | ✓ Executive or Full FDD report drafted from the room, as .pptx or Word | ✗ No report drafting | ~ Generic slides, not built from the whole data room |
| Excel workbooks | ✓ Multi-sheet .xlsx with live formulas, figures re-checked against source cells | ✗ Stores your models, doesn't build them | ~ Can build a sheet; no check against source cells |
| One Pager | ✓ One-page deal overview as a PDF | ✗ No deal summaries | ~ Summaries limited to the files you paste in |
| Red-flag memo | ✓ Ranked first-pass risks from statuses, conflicts, and open questions, as PDF or Word | ✗ No risk summary | ~ Can draft a memo, but not from the room's tracked statuses |
| QofE adjustments | ✓ Source-checked adjustments; only accepted rows reach adjusted EBITDA | ✗ Stores the numbers, no adjustments | ✗ No checked, reviewable adjustments schedule |
| Human review and sign-off | ✓ Banker or analyst accepts or rejects each finding, with comments and reasons | n/a No AI findings to review | ✗ Not purpose-built for the full DD workflow |
| Deal-level permissions | ✓ Enforced server-side, per deal room | ✓ Most offer some form of permissioning | ✗ Single shared account or session |
| Audit trail | ✓ Questions, views, uploads, exports, and decisions logged; export the log as CSV or Excel | ✓ Typically included for document access | ✗ No persistent deal-level audit trail |
Built for the people
doing the diligence.
AI-assisted, human-reviewed, at every step.
Investment Banking
First-pass analysis across large transaction rooms and faster evidence gathering.
Private Equity
Evaluate financial, commercial, legal, and operational evidence before investment committee review.
Transaction Services
Reduce repetitive document review while maintaining traceability.
Private Credit
Surface debt terms, covenants, collateral, and financial evidence faster.
Corporate Development
Accelerate acquisition diligence without losing source traceability.
Talk to us
Bring Vaultrix
to your deal team.
Tell us about your team and the deals you're working on, and we'll set up access and a plan that fits how you work. Someone from Vaultrix replies to every message.
Message sent.
Someone from Vaultrix will get back to you shortly at the email you provided.
Frequently asked questions
Answers for deal teams and compliance officers
PDF, Word (.docx), Excel (.xlsx/.xls), PowerPoint (.pptx), CSV, and plain text. Scanned, image-only PDF pages are run through OCR automatically, so their text is searchable and cited to the page like any native PDF. OCR runs only on pages that have no text layer, using Azure AI Document Intelligence within our Azure environment.
Yes. Upload the whole room as a single ZIP archive (up to 500 files) and Vaultrix parses, indexes, and categorises every supported file so the analyst can reason across all of it, not just one document at a time.
Every answer is required to cite the page, slide, or spreadsheet cell it came from. If the model can't point to a source, it says so instead of guessing. That doesn't make the model infallible; it means every claim is checkable. In Research mode, every claim is labeled as coming from the data room, a named web source, or general background knowledge.
Yes. Vaultrix parses .xlsx and .xls files sheet by sheet, and citations resolve to the exact sheet name and cell range, not just the filename.
Yes. When two source documents disagree on the same fact, Vaultrix surfaces both instead of silently picking one for you.
All drafted from the data room: an FDD report deck in PowerPoint (an Executive Report of about 10 to 12 slides, or a Full FDD Report of about 90 to 100 slides, also downloadable as Word), multi-sheet Excel workbooks with live formulas, a one-page deal overview as a PDF, a red-flag memo (PDF or Word), a quality-of-earnings adjustments schedule (Excel), a management question list (Word or Excel), and a request-list tracker (Excel). Exports carry your firm's branding. Every figure is traceable to its source inside Vaultrix; exported files are kept clean of citation markers. They are drafts for your team to review, not final sign-off.
Yes. Import your request list from Word or Excel and Vaultrix maps each line to evidence in the data room, proposing evidence found, partially supported, or no evidence found. It never closes a request on its own.
Yes, when you turn on Research mode. It's off by default and switched on per question. With it on, the AI Analyst can run live web searches for things the data room can't contain, such as peer benchmarks, recent news, or regulatory changes, while still preferring your documents for deal-specific facts. Web sources are listed separately from data-room citations, so you can always tell which is which. Search terms are derived from your question, so leave it off when the wording of a question is itself sensitive.
Document indexing runs entirely on our own infrastructure. The text of your files never leaves our servers for that step. Answering a question does call Anthropic's and OpenAI's APIs, which under their standard API terms don't train on your inputs or outputs by default. We haven't signed a separate zero-retention addendum yet. If that's a hard requirement, tell us before you upload anything sensitive.
Every deal is a separate access boundary enforced server-side. A user needs an explicit membership on a deal to query it, upload to it, or view its documents; org membership alone grants nothing.
Not yet. Today Vaultrix runs as a single, shared environment. Each org's data is isolated by role-based access control and org-scoped queries enforced in the application and database layer, not by separate infrastructure per customer. A dedicated deployment is realistic for a large enterprise engagement. Reach out and we'll scope it.
Vaultrix has not yet pursued formal SOC 2 or ISO 27001 certification. What we can tell you concretely: every deal enforces access control server-side, document indexing happens locally instead of through a third-party embeddings API, and every AI finding needs a human sign-off before it's treated as final. If a formal certification is a hard requirement today, let's talk before you sign up.
Your next data room
can be your first test.
Upload a transaction data room and see how Vaultrix maps evidence, surfaces gaps, and answers diligence questions with source-level citations.
No card required.