Coming soon

Know what your peers are doing. With the evidence attached.

Cohort Brief researches the organizations you choose against the questions you keep asking, and delivers briefings you can put straight into a leadership meeting. Every claim carries its sources, its evidence stage, and the date it was checked.

Built for leadership teams in higher education, healthcare, financial services, insurance, and public works.

Briefing Evidence stage: built Confidence: medium

Three of your seven peers now publish an approved-tool list; two tie it to procurement

What to notice

The lists were announced within four months of each other, and each names a responsible office. None yet reports usage or outcomes.

What it means for you

Your published policy predates all three. The gap is not the policy but the named owner and the list, which a peer will ask about at the next consortium meeting.

Sources (3, all read this cycle)
  • read Peer A, IT governance page, as of 2026-08-14
  • read Peer B, board minutes, as of 2026-07-02
  • read Peer C, procurement bulletin, as of 2026-09-09
An illustration of the artifact contract, not a real briefing: conclusion-led title, what to notice, what it means for you, counted sources, evidence stage, check date.
Why Cohort Brief

Peer intelligence usually arrives as opinion. This arrives as evidence.

Leadership teams already know which organizations they measure themselves against. What they lack is someone reading those organizations' public record every day and saying, with sources, what changed and why it matters here.

Your cohort, your questions

You name the organizations that matter and the questions you keep asking about them. A research template supplies the topic, such as AI adoption or market position, and starter questions you can edit, pause, or replace.

Evidence, not opinion

Every finding traces to a page the research team actually read. Figures are recorded from those pages with their as-of date, never lifted from prose. Claims are staged from announced to built to adopted to outcome, so you can tell a press release from a result.

Delivered where you decide

Briefings land in a workspace your whole team shares, and a daily brief email carries the new ones to whoever opts in. Ask the assistant a question and it answers from your workspace's own research, citing what it used.

How it works

A research cycle, run by a team of agents, held to rules a model cannot talk its way around.

Each cycle reads your cohort and your standing questions, then five named agents work in a fixed order. The harness, not the prompt, decides what they are allowed to claim.

  1. Set the lens Organization Context

    Reads your organization's profile and priorities so that everything downstream is interpreted for you, not for a generic reader. Your own organization is researched from public sources like any peer, and shown as "us".

  2. Search Landscape

    Searches the web for developments that answer your questions and proposes candidate findings as invisible drafts. It also follows threads from earlier cycles, so a finding grows rather than goes stale.

  3. Read Research

    Opens the pages, looks for primary sources, and attaches detail. A source counts only if it was fetched or returned by search in this cycle.

  4. Verify Verification

    Decides finding by finding whether the evidence supports publishing, assigns an evidence stage and confidence, confirms links between findings, and records any hard figures the sources state.

  5. Compose Synthesis

    Turns published findings into briefings and charts. It has no web access at all: it can only cite what Verification let through, and it can only chart recorded measurements.

  6. Interpret and deliver Organization Context, then the brief

    Writes what each new briefing means for you. Then the daily brief carries the new briefings, verbatim, to members who asked for them.

AnnouncedA plan, a press release, a stated intention.
BuiltSomething exists: a policy, a tool, a named office.
AdoptedPeople are using it, and there is evidence of that.
OutcomeA reported result, with a date and a source.
What you get

Four ways in, and a conversation that knows your research.

Members never need to learn the research taxonomy to use it. The workspace opens on what deserves attention and lets you go as deep as the evidence.

Now

What deserves attention

The newest briefings, ordered by what changed, with their evidence stage in view. Honest when nothing new arrived.

Explore

Ask by question

Six question-verb doors over the research: who is doing what, how far along, with what result. Every answer is a finding with its sources.

Landscape

Where you stand

Every organization, yours included, placed by breadth and depth of evidence. Positions are computed from findings, never judged by a model.

Library

What you kept

Briefings and charts you saved, with earlier versions preserved. Pin a chart to the dashboard and it follows the latest verified revision.

Chat and voice, grounded in your workspace

Ask in text or by voice. The assistant searches your workspace's findings and briefings, cites what it used, and can compose a new briefing or chart on demand, under the same evidence contract as the research team.

The daily brief

After each cycle, opted-in members receive the new briefings by email: the headline, what to notice, and the chart. Nothing is written for the email itself, and unverified briefings say so.

Sectors

One research method. Vocabulary, profile fields, and starter questions configured per sector.

A workspace is configured for its sector: the profile it keeps, the words it uses, the guidance its researchers follow, and the questions it starts with. The evidence rules are the same everywhere.

  • Higher education
  • Healthcare
  • Financial services
  • Insurance
  • Public works
Coming soon

Cohort Brief is in a closed beta with its first design partners.

If your leadership team keeps a list of organizations it watches, and keeps asking the same questions about them, we would like to talk.