Skip to content
MEDICIMEDICI home
Get early access

Sample report

All 11 sections. Nothing withheld.

The full structure of a MEDICI strategy report — the same fields the product emits, in the same order, with a fingerprint you can verify yourself.

The brief

Where should an AI strategy product sell first, and what would the first ninety days look like?

PE portfolio companies — AI strategy departmentBalanced mode
01

Executive Summary

Context synthesis and strategic positioning

Mid-market PE sponsors still buy strategy as a six-to-twelve-week consulting engagement. MEDICI compresses a first-pass strategy department into a single two-agent run: market exploration by Marco Polo, then an operator-grade growth plan by Michelangelo. The wedge is portfolio companies that cannot justify a full-time Head of Strategy.

02

Market Opportunities

Ranked by priority, each with a market-size estimate

high prioritylow-hanging fruit

Embedded strategy desk for PE operating partners

Sell a seat to the operating partner, not the portfolio-company CEO. Each new investment triggers a MEDICI briefing that becomes the 100-day plan appendix.

$1.8B annual PE/VC ops tooling spend in North America

PEoperating partner100-day plan
high prioritylow-hanging fruit

Founder briefings for seed and Series A

Founders need a board-ready strategy memo without an $80k consultant. MEDICI produces a fingerprinted report and a slide deck they can send to investors the same day.

40k US startups raising per year; $2k–$8k willingness to pay

foundersfundraisingdeck
medium priority

Corporate venturing scout memos

CVC teams run twenty or more scouting theses a year. A reusable agent pipeline with stored reports beats a graveyard of scattered documents.

Corporate venture units inside the Fortune 1000

CVCscouting
03

Growth Strategy

Go-to-market, target segments, key metrics and timeline

Land with ten design-partner PE operating partners in the lower-middle market, $200M to $2B AUM. Expand seat by seat across the portfolio. Founder self-serve is the secondary motion, opened once report and deck quality is proven.

Target segments

  • Lower-middle-market PE operating partners
  • Seed and Series A founders preparing a raise
  • CVC scouting teams

Key metrics

  • Briefings completed per week
  • Report-to-deck export rate
  • Paid seats per PE firm
  • Time to first report

Design partners in 30 days; paid pilots in 60; self-serve founder tier in 90.

04

Pricing Model

Tiers with features, and the rationale behind the structure

subscription

Studio

$100/mo

  • Unlimited briefings
  • PDF report and deck
  • 1 seat

Firm

$1,200/mo

  • 10 seats
  • Shared session library
  • Priority models

Platform

custom

  • SSO
  • Private model routing
  • Portfolio roll-up

Price against a junior strategy associate, not against a general-purpose chat assistant. The artifact — a fingerprinted report plus a board deck — is the unit of value.

05

Partnership Strategy

Who to partner with, and what the partnership is for

Integrate into the PE operating cadence through the deal-management systems these firms already run. Near term, partner with two boutique strategy firms who use MEDICI as a white-label first pass they then refine.

06

Compliance Notes

Regulatory and data-handling constraints on the plan

Do not ingest confidential deal memoranda without a data-processing agreement and a stated retention policy. Fingerprint every report so board packs are tamper-evident. Keep personal and patient data out of the briefing form entirely.

07

Engineering Directives

Technical recommendations for the build team

  • Keep the frontend and API on one origin so server-sent event streaming stays reliable.
  • Make the model name an environment variable; never pin a host-specific model identifier.
  • Gate the deployment until real authentication exists.
  • Persist the full structured report per session so later retrieval is not a data-backfill problem.
  • Record wall-clock duration and token cost per session from the first run.
08

30 / 60 / 90 Day Plan

Sequenced action items by time horizon

First 30 days

  • Stand up the local toolchain, Postgres and model access
  • Run ten internal briefings and keep the strongest as a reference
  • Put the deployment behind access control for investor links

By day 60

  • Sign five design-partner PE operating partners
  • Tighten the report schema against their redlines
  • Record a short product walkthrough

By day 90

  • Convert design partners to paid Firm-tier pilots
  • Open the founder self-serve waitlist
  • Decide between magic-link authentication and staying invite-only
09

MVP Features

Prioritized P0 to P2, with effort and rationale

  • p0small effort

    New briefing with a streamed two-agent run

    The moment that lands in a partner meeting

  • p0medium effort

    Strategy report and board deck export

    The artifact they forward internally

  • p1small effort

    Board Room dashboard

    Shows accumulated intelligence rather than a chat log

  • p1medium effort

    Session history and re-run

    Lets a thesis be refreshed as the market moves

10

Team Structure

Recommended roles with hiring timing and rationale

  • hire-now

    Founder / product

    Owns the briefing quality bar

  • hire-now

    Full-stack engineer

    Hosting, streaming, and the evaluation harness

  • 6-months

    Forward-deployed strategist

    Sits with operating partners on the first ten accounts

11

Resource Requirements

Personnel, infrastructure, tooling, monthly estimate and runway

Personnel
2 full-time equivalents for the MVP window
Infrastructure
Managed container hosting, managed Postgres, model access
Tooling
Standard TypeScript toolchain and a migration-managed database
Monthly estimate
$400–$1,200 infrastructure plus model spend at demo volume
Runway assumption
12 months to Firm-tier revenue

Report fingerprint — SHA-256

9fd7c180 92ba9f01 db3a5699 a73f7bdd f3b4fe28 84ab219b 0ae5d918 f601ea97

This hash is computed from the report’s content. Verify it yourself: take the JSON below and hash the file you receive.

shasum -a 256 medici-sample-report.json

Hash the file as delivered rather than a copy you have re-saved. The fingerprint covers the exact bytes, so reformatting the JSON — even just adding indentation — produces a different hash. That is the property that makes it tamper-evident.

What this does not do: it does not prove the analysis is correct, it is not a timestamp, and it is not a legal attestation. It proves integrity, not truth.