THE THESIS

Venture building,
paid in tokens.

LFG funds founders with AI agent work instead of cash, for equity or a revenue share. Tokens measure what we put in and what comes back.

Tokensthe unit of investment
Tranchesreleased milestone by milestone
One stackshared by every venture

Why now

01

Ideas are everywhere

From a plumber's booking tool to an AI observability platform.

02

Agents can do the work

Product, marketing, support and operations, end to end.

03

Founders bring what's scarce

The insight, the network and the selling.

04

Tokens make it measurable

Every tranche has a cost, a plan and a result.

The loop

One cycle per tranche. Results decide the next one.

  1. Assess

    The analyst interviews the founder and checks public evidence.

  2. Fund

    A tranche sized to the next milestone. Both sides sign.

  3. Build and run

    Agents ship and operate. The founder sells.

  4. Measure

    Revenue, usage, spend and founder commitments.

  5. Decide

    Release more, hold and learn, or halt.

Repeats at every milestone or weekly check-in

How the analyst sizes a tranche

Fund the next milestone, not the whole company. Only as much as the expected return justifies.

What it weighs

  • MarketBottom-up TAM, SAM, SOM in US dollars
  • Founder fitExperience, access, network, capacity
  • EvidenceSourced facts vs self-reported claims
  • Next milestoneWhat must be built to learn the most
The rule

Tranche = cost of the next milestone

only if expected return ≫ tranche

FundNeeds more infoDeclineProhibited

Every claim is labelled sourced, founder-reported or estimate. Unknowns stay unknown.

100K–1M tokensLanding page + content/distribution strategy
1M–5M tokensSmall internal tool or no-code-backed MVP
5M–50M tokensProduction SaaS MVP with auth, payments and integrations
50M–500M tokens, stagedComplex platform (e.g. observability, data infrastructure)
Worked exampleBooking app for independent plumbersFounder: a working plumber in 3 WhatsApp groups with ~400 peers
MVPWhatsApp booking + calendar2.2M
Landing pageSign-up and pilot deposits0.5M
SupportFAQ bot for customers0.5M
First tranche3.2M tokens
Deal8% of revenue until 3× the tranche, or 36 monthsFounder commitmentsInterview 10 plumbers · 3 paid pilots · demo video in the groups

Agents do the work, on one stack

Each venture gets a team of agents and a production platform from day one.

Build

MVPs, features, landing pages, integrations

Grow

Content, campaigns, outreach, sales support

Run

Customer support, bookkeeping, analytics

Maintain

Hosting, upgrades, security fixes, uptime

Shared core, built once
AuthPaymentsAnalyticsEmailSupportHostingSecurity

Every new venture starts production-grade and needs fewer tokens to launch.

Release, hold or halt

At each check-in the analyst reads real signals and recommends; LFG decides.

RevenueActive usersConversionTokens spent vs planWork shippedFounder commitments

Release

Milestone hit, numbers moving.

  • Next tranche, sized to the next milestone
  • Next terms reflect the traction

Hold and learn

Mixed or missing signals.

  • Small, targeted experiments
  • Founder gathers specific data
  • Reassess at the next check-in

Halt

No traction after real tries, or risk.

  • Spending stops
  • The founder keeps the venture and its customers
  • Learnings feed the next decision

How LFG earns

  • EquityFor venture-scale companies that will raise capital.
  • Revenue or profit shareFor smaller businesses. Capped at a multiple of the tranche, with an end date.
  • Return per tokenTokens in, revenue back. Comparable across every venture.

Why it compounds

  • Cheaper launchesEach shared service makes the next venture cost fewer tokens.
  • Sharper decisionsEvery decision and its outcome calibrate the analyst.
  • More shots on goalSmall tranches let LFG back many founders and double down on winners.

The analyst decides. Agents build. Results earn the next tranche.

See it from the founder's side.