If you’ve ever opened your inbox at 6am and wished someone had already filtered to what matters — you already know what we built.

Researcher is the daily-intelligence analyst your team would hire if they could find one. We trained it on the way principal-class operators actually read.


01 · The surfaces

Three surfaces. One operating loop.

SurfaceWhat it doesCadence
Daily briefScores 50–100 candidate items overnight against your evolving signal map. Surfaces five things worth your morning attention, with context.Every morning, before your first meeting
Slack chatAsk anything. Researcher reshapes weak questions into sharper ones, then answers from your substrate — coaching the way you think, not just the way you skim.On-demand, real-time
Weekly meta-briefThe honest read across the week. Patterns you didn’t notice. Trends your reactions are telling us. What the week was really about.Sundays
reactions compound — the next brief is sharperDaily briefFive things each morningYou reactLabel what mattersSubstrate sharpensYour signal map compoundsTHREE SURFACES · ONE COMPOUNDING SUBSTRATE

Memo register. Not a chatbot pretending to be a coworker — an analyst that learns the way you actually read.


02 · The wisdom-tier underneath

Most “AI for research” tools are knowledge layers — they retrieve facts. Researcher is built on three.

KnowledgeSources · retrieves · deduplicates — table stakesUnderstandingSignal vs. noise — trained on your reactionsWisdomWhat to surface, hold, bury, or flag — months of judgment, codifiedTHE MOAT
LayerWhat it doesWhy it matters
KnowledgeSources, retrieves, deduplicates across your subscriptions and the open web.Table stakes. Every tool does this.
UnderstandingDiscriminates: what’s signal for you, what’s noise, what’s a pattern you’ve already seen.Trained on your reaction history — your labels become a learning corpus.
WisdomApplies judgment: what to surface now, what to hold, what to bury, what to flag as “you should know this even though you usually skip it.”This is the moat. Months of your reading habit, codified.

The Proverbs distinction — knowledge → understanding → wisdom — is our internal IP frame. We build every productized role to operate at the wisdom tier.

Four-layer system-prompt architecture. Substrate truth-check before every brief. Reaction-loop learning between cycles. No hallucination on your own corpus.


03 · The reader

Who Researcher is for.

Operator-class principals running their own intelligence layer:

Industry strategists

Track 40+ sources daily, can’t keep up, and want their pattern-recognition compounded.

Specialist principals

Legal, medical, financial advisory — need domain signal filtered without ceding judgment to a generalist tool.

Asset-management principals

Run multi-property or multi-portfolio operations and want intel without building an internal team.

Founder-CEOs

Need a research analyst at $1,500/mo, not a $200K/yr hire.

Five named warm-leads pre-sold Researcher to their networks before it shipped. That’s the customer pattern: principal-class operators selling it forward, because the alternative — building this internally — is structurally harder than buying it.


04 · Setup

What setup actually looks like.

Reverse-Sherpa pattern. We build the mountain. You ratify.

PhaseWhat happensYour time
Week 1We map your reading habit — sources you subscribe to, threads you track, signal you react to — and stand up your tenant substrate.90 minutes of setup conversation
Week 2Daily briefs start firing. You react with labels — 📝 for “synthesize this further,” 👀 for “track this thread.” The substrate learns.5–10 minutes / morning
Week 3+The wisdom tier compounds. Reaction history shapes the signal map. Meta-briefs surface patterns you didn’t notice. Slack chat goes from useful to load-bearing.What you spend on intelligence already — compounded

Key ownership

Founding cohort runs on our Anthropic key — we absorb the LLM cost, you absorb zero infrastructure setup. When you scale beyond founding-cohort terms (or your team requests procurement-controlled billing), we migrate you to a customer-owned-key pattern. Your stack, your trust boundary.


05 · Founding cohort pricing

$1,500 / month

Locked life-of-engagement.

All three surfaces (daily / chat / weekly meta-brief)No surface gating, no usage caps within reason.
All LLM costs (founding cohort)We hold the Anthropic key; you pay $1,500 flat.
Direct access to the team that built itMeridian Intelligence is one operator and an AI org chart. You talk to us.
Case-study rightsWe name you publicly when you say yes; never if you say no.

Researcher runs at 93%+ gross margin at $1,500/mo across every observed usage tier. Founding-cohort pricing is a signal of friendliness, not a cost-recovery floor. We price for relationship velocity, not COGS.

Cohort size: 5–10 clients. Closed when full.


06 · The trajectory

Researcher is the first productized role from Meridian Intelligence. Drafter, Biz Dev, and full AI Org Chart configurations are in deployment with other founding clients. The Researcher-MCP server is on the Q4 2026 roadmap — your team’s existing tools (Claude, ChatGPT, Cursor) will engage Researcher as a peer agent through MCP, not as another dashboard you context-switch into. The substrate compounds; you don’t churn into yet another tool.

No form. No scheduling link. Just a reply.

Email us with a sentence about what you read every morning and what you wish were already filtered. We’ll reply with whether founding cohort is a fit — and what week 1 would look like for you.

founding-cohort@meridian-intelligence.io

If the timing isn’t right, no issue. If it is, the door is open.