§ 00Intro

Find the people who know what the internet doesn’t.

Vesper turns high-stakes research questions into ranked maps of reachable domain experts — then helps teams compound every conversation into institutional memory the public web cannot reach.

Request access See the console
WedgeExpert sourcing for diligence
Surface200k+ reachable operators
Coverage14 verticals
StatusPrivate beta
§ 01The shift

Public knowledge is no longer proprietary.

Every investor can ask the same model to read the same filings, transcripts, reviews, and market reports. The edge has moved upstream — to humans with firsthand experience, buried context, and judgment earned inside the market.

A · The old waycirca 2024

Expert access still runs like a services business.

Analysts translate a thesis into expert profiles. Teams manually search LinkedIn, industry sites, and databases. Contact paths are enriched one by one.

  1. Translate thesis into a list of profiles, by hand
  2. Search LinkedIn, Google, niche directories, archives
  3. Enrich contact paths one row at a time
  4. Run outreach from spreadsheets and inboxes
  5. Conduct calls, take scattered notes
  6. Knowledge disappears the moment the project ends
B · The Vesper way2026 —

Primary research becomes software.

A research question becomes a structured operation: source the right people, surface why they matter, reach them, capture the conversation, and reuse the insight forever.

“AI can summarize the internet. Conviction still comes from people who lived the problem.”
— Vesper, founding memo
§ 02Demonstration

Pick a brief. Watch the map change.

The console turns a vague diligence need into ranked candidates, relevance signals, and a reason to start the conversation.

Vesper / Console v0.4Brief Workspace · Diligence ModeLive index
Active briefs · 03
Why do inventory automation pilots stall before regional rollout in specialty retail?
38 matches
Pilot historyFrontline adoptionRegional rollout
№ 01
Former VP, Store Operations
National specialty retailer · 600+ locations
Led labor planning and inventory accuracy programs across more than 600 locations. Owns the lived friction between pilot success and chain-wide rollout.
Industry talk + operating history
94
Fit score
№ 02
Ex-Director, Fulfillment Strategy
Regional grocery network · 9 states
Owned store-to-door rollout planning, vendor evaluation, and exception handling workflows during a stalled automation deployment.
Trade panel + public profile
91
Fit score
№ 03
Independent Retail Systems Advisor
Former inventory platform operator
Advises operators on pilot design, frontline adoption, and multi-site change management for retail technology.
Advisory profile
87
Fit score
§ 03The system

A system of record for human intelligence.

Ask a research question once. Build an expert map, contact strategy, call trail, synthesis layer, and institutional memory around it — in the same place.

№ 01
Source
Expert sourcing engine
Turn a thesis into a ranked map of operators, customers, suppliers, and former executives — with a defensible reason for every name on the list.
200k+indexed
№ 02
Enrich
Deep enrichment
See why each person matters, where the signal came from, and the best available contact path. Sources surfaced, not buried.
14signal types
№ 03
Reach
Assisted outreach
Coordinate the human workflow without making AI phone agents the product. Drafts, threading, and reply-state, all in one ledger.
3.4×response lift
№ 04
Capture
Knowledge capture
Transform calls and notes into reusable institutional memory. Searchable, citable, attributable across every future project.
∞half-life
Operating sequence · 07 stagesLive
01Ask a question
02Source experts
03Enrich paths
04Route outreach
05Capture calls
06Synthesize
07Compound memory
§ 03·BBuilt for
iInvestment teams
iiFamily offices
iiiPrivate equity
ivConsultants
vCorporate strategy
viOperators
§ 04Vision

From expert networks to the Human API.

A living map of human expertise — retrievable, reachable, and reusable.

The next generation of research software will not just retrieve documents. It will know when to ask a human, who to ask, how to reach them, and how to turn the answer into reusable institutional memory.

§ 04·APillars
i.
Human graph
Retired operators, former executives, customers, suppliers, pilot leads, procurement heads, and domain specialists — structured, searchable, defensible.
ii.
Decision layer
A workflow for diligence, vendor validation, market research, and operational questions — not a chat box bolted onto a CRM.
iii.
Compounding memory
Every permissioned conversation makes the institution smarter for the next project. Knowledge stops disappearing when projects end.
§ 05Request
access

Build conviction from knowledge your competitors can’t scrape.

We review every request
by hand.