Skip to content
Due Diligence

The investment memo, sourced to the page.

Upload the data room. Eight specialised agents ingest your files, cross-check the public record, and return a structured DD memo with file-anchored citations. Built for M&A and PE associates who answer to an IC on Monday.

No credit card. One DD run is the 59-credit bundle. 50 free credits cover a partial run; top up from $9.

Used by associates at
Why DD teams switch

Three problems that disappear.

The problem

The data room is a maze and the deadline is Monday.

CIM, model, contracts, customer references, market sizing — an associate spends the first three days just orienting. Half the memo writes itself if you can get there fast enough.

With Zyphv

Drop the room in. Get a structured first draft.

Files upload via a shared pipeline; document agents extract, chunk, and embed. The Researcher cross-references public sources. You receive a structured memo with citations into your own files.

The problem

An LLM that hallucinates a customer kills the deal.

An associate cannot put a number in the IC memo unless they can show where it came from. A confident summary without sources is worse than no summary.

With Zyphv

Every claim cites a file or a URL.

The DD pipeline anchors each statement to a specific document chunk or a public source. Provenance hashes per agent let you audit the chain end to end.

The problem

Generic AI tools don't speak diligence.

Most "AI for deals" tools are wrappers around a single model. They cannot sequence document review, market validation, financial sanity checks, and risk flags in a single pass.

With Zyphv

Eight agents with specific jobs.

Each agent owns one section: market, financials, customers, team, risks, comparables, integration, summary. The Compiler assembles a memo your MD can read.

How it works

Eight agents. One memo.

Files in. Sourced DD memo out. Each step writes to the provenance log so you can defend every number.

01
Document Ingest
Parses CIM, model, contracts, decks. Token-based chunking, vector index per run.
pgvector · embeddings
02
Market Researcher
Validates market size, growth, and competitive set against the public record.
grok-3 · live search
03
Financial Analyst
Reads the model, flags unit economics, working-capital, and revenue-quality concerns.
claude-sonnet
04
Customer Analyst
Customer concentration, churn, contract terms, references where available.
claude-sonnet
05
Team & Org
Leadership track record, hiring velocity, retention signals.
claude-sonnet
06
Risk Mapper
Legal, regulatory, technology, and customer-concentration risks; severity scored.
claude-sonnet
07
Comparables
Trading and transaction comps with the rationale spelled out.
grok-3 · web
08
Compiler
Polished IC-ready memo with executive summary, structured sections, and citations.
claude-opus
Sample output

What an IC-ready memo looks like.

Rendered in-app and exportable to PDF. Below is a representative excerpt.

Due Diligence — Project Halcyon (SaaS, mid-market)
Run dd-92c · 8 agents · 59 credits · 18 files · 41 sources
2026-06-04

Quality of revenue is strong; customer concentration is the active risk.

Halcyon shows 142% net revenue retention across 87 customers, but the top three account for 38% of ARR — two on renewals within 9 months. Underwriting should price renewal risk and verify multi-year commitments.

142%
NRR (LTM)
38%
Top-3 customer share
9 mo
Concentration risk window

High: Customer concentration as above. Medium: Pricing model assumes 7% net price uplift in FY27; competitor pricing data does not support it. Low: Two key engineers without retention agreements.

CIM p.14FY25 model · ARR tabCustomer list (data room)g2.comlinkedin.comSEC filings · competitor
Pricing

Credits, not seats. Use them on any swarm.

The DD bundle is 59 credits per run. Sign up for 50 free credits, then top up as needed.

Starter
$9 / 100 credits

~1 DD run plus follow-ups. Best for piloting on a real target.

  • Full eight-agent DD pipeline
  • Data-room file upload
  • PDF + web memo export
  • Provenance + file-anchored citations
Choose Starter
Scale
$69 / 1,000 credits

~16 DD runs. For diligence pods and deal teams.

  • Everything in Professional
  • Cross-run intelligence
  • API + webhook integration
  • Shared workspace
Choose Scale

DD run: 59 credits (bundle). File processing is included in the bundle.

We don't promise magic. We deliver clarity when it matters most.

Questions

Reasonable answers.

How long does a DD run take?
DD runs typically complete in 10–20 minutes depending on data-room size. You can watch progress live; the report waits for you if you step away.
What file types can I upload?
PDF, DOCX, XLSX, PPTX, CSV, TXT. Files are extracted, chunked into ~500-token segments, embedded into a per-run vector index, and tied back into citations.
Is the data room safe?
Yes. Uploaded files are tenant-isolated, never shared across tenants, and never used to train models. You can delete files at any time and they are removed from the vector index.
Can I cite specific pages in the memo?
Yes. Each section in the output viewer carries source references that link back to the specific document chunk or public URL the agent used.
What if a run fails?
If the pipeline cannot complete, credits are refunded automatically. If you're unhappy with quality, email us — every message is read.
Do you support custom DD checklists?
The Studio + Marketplace let you publish your own agents and slot them into the pipeline. Reach out if you want a guided custom playbook.
More from Zyphv

Explore other intelligence verticals.

Ready when you are.

Start free — 50 credits on us. No card. No noise.

Begin your first DD run