Evidence was fragmented
Official sites, filings, allocation documents, manager pages, PDFs, news, and internal records each contained only part of the investor picture.
All case studies
Private marketsInvestor finder
Institutional investor qualification and relationship intelligence for private markets
A searchable investor database with real institutional names, type and geography filters, source-backed profile fields, relationship context, and visible evidence gaps.
Real institutions
searchable profiles
Evidence-linked
recorded claims
Relationships
fit and context
Client context
An institutional investment team working across LP, GP, strategic, and relationship intelligence.
The customer is anonymous. Public institutional facts are linked to their source; private relationship notes and targeting decisions are not displayed.
Evidence boundary
Sanitized product patternLP research workflow notesPublic institution source records
The operating problem
A list of institutions is not investor intelligence.
The team could find names. The harder task was establishing which entity was the right one, what its mandate and relationships actually were, which claims had evidence, and what still had to be verified before a briefing or outreach decision.
Identity errors compounded downstream
Aliases, similarly named institutions, managers, funds, and subsidiaries made a weak match dangerous to normalize into a canonical profile.
Research decayed after the project
One-off spreadsheets preserved conclusions but rarely preserved source lineage, conflicts, freshness, and the next verification step.
The product thesis
A maintained evidence pipeline for investor decisions.
The product treats discovery as the beginning—not the answer. Candidate sources pass through relevance and identity checks, structured facts retain provenance, conflicts remain visible, and deterministic profiles and briefings sit above the evidence layer.
Gate before writing
Candidate evidence is checked for entity fit, relevance, duplication, and source strength before it can update a maintained record.
Preserve claim-level lineage
Mandate, allocation, portfolio, relationship, and contact signals remain connected to the source from which they were derived.
Make gaps part of the product
Missing, stale, conflicting, or review-pending fields remain visible so an analyst knows what to verify next.
Product demonstration
See the workflow move from approved input to reviewed output.
The interactive view shows what the system does at each stage, which evidence moves forward, and where professional review enters the process.
Investor finder
Search and qualify institutional investors
1Define the mandate
2Resolve identity
3Check the evidence
4Shortlist or reject
Candidate intake and entity resolution
Discovery is not a profile. A candidate must resolve to one entity, show usable evidence, and pass mandate review before it reaches a shortlist.
Decision trace
Merge this evidence with the existing profile; do not create a second record.
Selected investor
Canada Pension Plan Investment Board
Public pension · Canada
Net assets
CAD 793.3bn at FY2026 year-end
Combined Fund return
7.8% for FY2026
Scope control
Fund return kept separate from 13.2% benchmark
Decision-ready output
Build a briefing only after the evidence and mandate checks are visible.
Public institutions and dated facts remain linked to official sources. Private relationship notes and targeting decisions stay access-controlled.
End-to-end workflow
From first input to a controlled next action.
The system discovers candidate evidence, checks identity and relevance, extracts structured signals, preserves source lineage, and serves profiles, briefings, feeds, and analyst answers from a durable evidence layer.
01 · Discover
Find candidate institutions and evidence
Search, official sites, filings, PDFs, feeds, and reviewed internal records create a candidate evidence set.
02 · Gate
Resolve identity, relevance, and duplicates
The system checks whether evidence belongs to the target and whether the entity already exists under another name.
03 · Structure
Extract and normalize decision signals
Accepted evidence becomes profiles, mandate context, portfolio observations, relationships, and freshness signals.
04 · Decide
Serve a briefing with evidence and gaps
A professional can screen fit, open the supporting source, see unresolved questions, and prepare the next verification or outreach action.
What was built
From discovery engine to relationship briefing.
The product is a durable research system, not a chat window sitting above an unexamined directory.
Search
Entity and evidence discovery
Institution, manager, fund, document, contact, market, and relationship discovery across defined source classes.
Evidence
Extraction and provenance
Structured observations from accepted documents and pages, retained with source and association context.
Profile
Canonical investor record
Normalized identity, mandate, portfolio, allocation, relationship, readiness, freshness, and conflict signals.
Brief
Decision and outreach preparation
Source-backed briefings, relationship maps, target context, and the specific checks still required.
The evidence layer can disagree with itself—and the product must show that.
Uncertain identity matches, conflicting observations, weak sources, and potentially stale facts enter review rather than silently becoming headline truth. Known-entity reads use maintained profile data before open-ended generation.
Canonical identity checks
Claim-level source traceability
Review queues for uncertain evidence
Visible freshness and readiness signals
Operating model
The change is in how the work begins, moves, and gets reviewed.
Before
Research ended in static lists and disconnected notes.
Source quality and entity identity were difficult to audit later.
Analysts repeated discovery whenever a new question arose.
After
Profiles are built from retained, associated evidence.
Conflicts, gaps, and freshness are visible beside the investor view.
Briefings begin from maintained context and identify the next check.
Reviewable outputs
Investor profilesRelationship mapsSource-backed briefsTargeting context
Operational change
Investor research became a maintained evidence workflow rather than a one-off search or an untraceable spreadsheet of names.
Implementation pattern
Start with the real operating boundary.
01
Define the decision schema
Specify target entity types, mandate signals, relationship types, source hierarchy, and what qualifies as a usable profile.
02
Build the evidence path
Connect discovery, gating, extraction, normalization, review, and canonical profile reads in that order.
03
Add decision surfaces
Expose search, profiles, relationship views, evidence review, briefings, and scheduled refresh only after the underlying record is traceable.
Approved inputs
Institutional websites and filingsFund and allocation documentsRelationship and portfolio contextReviewed internal records
Where this transfers
Relevant to fundraising, institutional coverage, strategic-capital mapping, manager research, and private-market relationship teams.
The takeaway