Five layers of tooling. One missing layer that keeps the decision alive.
A practical map of the PE software stack — and the decision integrity layer that has to sit above all of it.
Most private equity firms run on a stack: data and intelligence, sourcing, CRM and workflow, AI analyst tools, fund admin, portfolio reporting. The issue is not whether those tools work. The issue is that none of them owns the IC decision after the deal closes.
The typical PE workflow looks like this: opportunity found in sourcing tools, context gathered in data platforms, workflow managed in CRM, analysis accelerated in AI tools, reporting rebuilt manually in slides and memos. Five handoffs. The IC decision lives in the gap between layer four and layer five — and nothing in the stack carries it forward.
The stack, layer by layer
1. Market data and company intelligence
Examples: PitchBook, Capital IQ, PrivCo, CB Insights.
What they do: comps, market context, company reference data.
Where they stop: market context does not survive at IC. The data informs the decision; it does not defend it later.
2. Sourcing and relationship systems
Examples: Grata, Axial, Affinity.
What they do: target discovery, sourcing workflow, relationship coverage.
Where they stop: sourcing finds the deal. It does not own the underwriting record once the team commits.
3. CRM and deal workflow
Examples: DealCloud, Dealpath, Altvia.
What they do: process management, approvals, pipeline visibility, task coordination.
Where they stop: process is managed but the IC memo, the risk signals, and the operator-update drift live elsewhere — usually in spreadsheets, slide decks, and shared drives.
4. AI analyst and modeling tools
Examples: BlueFlame, Endex.ai, Microsoft Copilot, Macabacus, Hebbia.
What they do: accelerate analyst tasks, summarize documents, draft memos, support modeling.
Where they stop: faster work, but the output is unstructured artifacts on a shared drive. Speed without survivability is faster confusion.
5. Fund administration and LP operations
Examples: Dynamo, Juniper Square, Allvue.
What they do: investor reporting, fund administration, capital operations.
Where they stop: the LP report publishes the numbers. It does not own the underwriting that produced those numbers, or the operator update that changed them.
6. Decision integrity layer
Example: Capital Refinery.
What it does: takes the output of the layers above and turns it into a graded IC record that survives the deal — lane-aware IC memo, Risk Signals scoreboard, source-backed numbers, IC anchor on approval, drift tracking after the room votes, provenance-signed exports for LP and audit.
This is the layer most stacks are missing. It is the layer that answers “why is this asset still held?” two years after close.
What each layer does — and where decision integrity has to come from.
| Capability | Layer · Tools | Capital Refinery |
|---|---|---|
| Market data | PitchBook, Capital IQ, PrivCo | Context — does not survive at IC |
| Sourcing | Grata, Axial, Affinity | Finds the deal — does not own underwriting |
| CRM and workflow | DealCloud, Dealpath, Altvia | Manages process — IC record lives elsewhere |
| AI analyst tools | BlueFlame, Endex, Copilot, Hebbia | Faster drafts — output is unstructured |
| Fund admin / LP ops | Dynamo, Juniper Square, Allvue | Reports the numbers — does not defend them |
| Decision integrity layer | Capital Refinery | IC memo, Risk Signals, drift tracking, signed exports |
Where the stack actually breaks
Common failures we see in production stacks:
- underwriting assumptions are not preserved after the IC memo is uploaded to the shared drive
- downside scenarios live in isolated spreadsheets that age out within a quarter
- post-close monitoring tracks today's KPIs but never reconnects them to the entry assumptions
- committee outputs are rebuilt from scratch every quarter instead of generated from a live record
- by the time a covenant breach is visible to the team, the time to act has already passed
Why this matters now
$3.6 trillion of unrealized value sits on GP balance sheets (Bain 2025). Every one of those positions has an IC memo somewhere. Most of them have aged out. The team that can rebuild the decision basis on demand wins the LP question, the audit, and the next fundraise.
Map your stack to the layer that owns the IC decision.
Bring us your stack and a deal you closed last quarter. Same-day diagnostic pack back showing exactly where the decision basis broke and what it costs you.