Persistent Deal Intelligence

Context that never breaks.
Intelligence that compounds.

Vetrn builds a persistent, queryable intelligence layer for every deal — at both the deal level and across your entire organisation. Nothing gets lost. Every interaction makes it richer.

Most tools forget
everything between sessions.

Context amnesia is the silent failure mode of AI in investment workflows. Here’s what it looks like — and what Vetrn does instead.

Without Persistent Memory

Starting from scratch — every time.

  • Re-explaining the deal context every time you open a new session
  • Call notes from last week aren’t connected to the deck analysis from two weeks ago
  • A new analyst joins the deal and has no access to what’s already been researched
  • You ask “what was the CAC figure the founder mentioned?” and nobody can find it
  • Inconsistencies caught in a call aren’t surfaced when you prepare for the next one
  • Data room documents sit in a separate folder, disconnected from research outputs
  • IC prep means manually pulling information from five different places
With Vetrn Memory

Full context — always intact, always connected.

  • Every interaction draws on the complete history of the deal automatically
  • Call transcripts, research outputs, data room docs — all integrated and queryable
  • New team members get instant access to everything that’s been researched and verified
  • Ask any question about any detail from any point in the deal — get an accurate answer
  • Inconsistencies flagged in calls are remembered and surfaced in all future interactions
  • Every document in the data room becomes part of the searchable intelligence layer
  • IC prep draws on a single, complete, structured knowledge base

Memory at the deal level.
Memory at the firm level.

Our memory layers operate at two scopes — every deal gets its own intelligence layer, and your organisation accumulates knowledge across all of them.

Deal-Level Memory

Everything about a deal — unified and queryable.

From the first deck upload to the final IC discussion, every piece of information about a deal is integrated into a single persistent intelligence layer. All of it queryable, all of it contextually connected, all of it retained across sessions and team members.

  • Pitch deck content and thesis-fit analysis
  • Vedi company intelligence report
  • Vetra sector intelligence report
  • Full data room — financials, legal, contracts, patents
  • Meeting transcripts and structured post-call summaries
  • Flagged inconsistencies and unverified claims log
  • Analyst notes, questions, and commentary
  • Version history — every change and update tracked
Organisation-Level Memory

Intelligence that builds across your portfolio.

Vetrn builds institutional knowledge at the firm level — pattern recognition across deals, sectors, and founders that accumulates over time. The more you use it, the sharper the intelligence becomes across every new deal you evaluate.

  • Cross-deal pattern recognition — what signals predict outcomes in your portfolio
  • Sector intelligence that compounds over multiple analyses
  • Portfolio company comparisons available on any deal
  • Thesis evolution tracked — how your investment criteria shift over time
  • Team knowledge shared automatically — no institutional knowledge leaving with an analyst
  • Historical deal performance correlated to early-stage signals
  • Fund-wide research standards maintained and updated

Powered by Vetrn Memory

Vetrn Memory is the persistent memory infrastructure that makes Vetrn’s intelligence layer possible. Unlike session-based AI tools that reset between conversations, Vetrn Memory maintains a structured, long-term knowledge graph at both the deal and organisation level. Every interaction — a question asked, a document uploaded, a claim flagged — is integrated and retained. This is what makes Vetrn feel less like a tool you use and more like a team member who was in every meeting.

The full timeline of a deal —
all of it retained.

Every meaningful event in the life of a deal is captured in the memory layer. Here’s what that looks like from first contact to IC.

Day 1

Pitch deck ingested — thesis fit scored

The deck is parsed, signals extracted, and thesis fit scored with explicit reasoning. This initial analysis becomes the foundation of the deal’s memory layer — all subsequent intelligence is layered on top of it.

Deck signalsThesis fit scoreGap analysisClarifying questions
Day 2–3

Vedi & Vetra analysis completed

Both research engines run and deliver their outputs. All findings — verified, unverified, and flagged — are integrated into the deal memory. The intelligence layer now has a complete, structured picture of the company and its sector.

Company intelligenceSector intelligenceVerified claimsRisk signals
Day 5

Data room documents uploaded

Financials, cap table, legal docs, and contracts are added. Every document is parsed and its signals integrated into the memory layer — immediately queryable through the deal chat.

FinancialsCap tableLegal documentsContracts
Day 7

Founder call — Meeting Copilot active

The call transcript, structured summary, flagged inconsistencies, and open research checklist are all added to the memory layer. Every claim made in the call is cross-referenced with what’s already in memory — contradictions are flagged immediately.

Call transcriptPost-call summaryInconsistenciesOpen items
Day 12

Second call — follow-up questions prepared from memory

The Meeting Copilot draws on the full memory layer — the first call transcript, the research outputs, the unverified claims — to generate targeted follow-up questions for the second meeting. Nothing falls through the cracks.

Follow-up questionsUnresolved claimsNew inconsistencies
Day 15

IC memo generated from the complete memory layer

The investment memo draws on everything in the deal’s memory — all research, all call transcripts, all data room analysis, all analyst notes. A complete, citation-backed IC document exported in minutes.

Full memo exportCitation-backedIC-formattedComplete audit trail

Ask anything about a deal.
Get answers with full context.

Every deal has a dedicated AI chat interface that draws on the complete memory layer. It knows everything — and it never forgets.

Analyst
What was the CAC figure the founder mentioned, and does it match what we found in Vedi?
Vetrn
In the 14 March call, the founder stated a blended CAC of $340. Vedi’s analysis — based on publicly available cohort data from their LinkedIn ad spend and estimated conversion rates — put the figure at $480–$520. This is flagged as an inconsistency in the deal record. The gap is approximately 35–50%. Worth pressing on in the next call, particularly around whether the $340 figure includes all customer acquisition cost centres or only paid channels.
Analyst
What are the top three unresolved items before IC?
Vetrn
Based on everything in the deal record: (1) CAC discrepancy — founder’s figure vs. Vedi’s estimate is unresolved. (2) The Series A cap table shows a 22% employee option pool not mentioned in the deck or either call — legal has been flagged but not yet explained. (3) Vetra noted significant regulatory movement in the EU on data localisation that could impact the company’s roadmap — the founder hasn’t addressed it and it hasn’t been raised.
All answers draw on the complete deal memory layer — deck, research, data room, call transcripts, and analyst notes.

The intelligence that builds
across your entire fund.

Over time, Vetrn becomes a repository of your fund’s institutional knowledge — patterns, signals, and sector intelligence that improve every decision you make.

Cross-deal pattern recognition

As your deal history grows, Vetrn surfaces patterns across deals — signals that tend to correlate with strong outcomes vs. weak ones in your specific portfolio context.

Institutional knowledge that stays

When an analyst leaves, their knowledge of every deal they touched stays in the system. New team members get up to speed in minutes, not months.

Sector intelligence that compounds

Every Vetra run adds to your fund’s understanding of the sectors you focus on. Intelligence compounds — so your third analysis of a fintech deal is sharper than your first.

Portfolio comparisons on demand

Ask how a new deal compares to portfolio companies on any dimension — unit economics, market position, team profile, growth trajectory.

Thesis evolution tracked

Your investment criteria evolves. Vetrn tracks how your thesis changes over time — and can tell you how deals evaluated six months ago would score against your current mandate.

Standardised quality at scale

Organisation-level memory means every analyst — junior or senior — works with the same knowledge base. Research quality stops being a function of individual seniority.

See what a deal with full
memory looks like.

Book a demo and we’ll walk through a live deal — showing exactly how the Vetrn Memory layer connects every piece of intelligence from first look to IC.

Book a DemoFor Founders →