Skip to content

Likeness — Intake Brief

Phase: 1 — Intake Project: likeness Date: 2026-05-09 Confidence: High (synthesized from existing planning artifacts plus founder Q&A)

Note on this brief

This intake brief is synthesized rather than interview-derived. The founder has already produced a thorough planning corpus at the repo root (docs/founder-brief.md, docs/budget.md, docs/cofounder-jds.md, docs/ml-lead-technical-brief.md). Re-asking intake questions from scratch would have been wasteful. What follows compresses those documents into the skill's standard intake shape and incorporates three Q&A answers that materially affect downstream phases.

If anything below contradicts the source documents, the source documents win — flag it and I'll reconcile.

The idea

Likeness is a subscription platform where verified adult creators monetize controlled access to their own AI likeness. Creators upload real photos and videos as on OnlyFans, and can opt in to having a private AI model trained on the platform. Subscribers use that model to generate images within rules the creator sets, submit favorites for review, and access creator-approved galleries. Every output is watermarked, tied to a license, and revocable. Models never leave the platform.

One-line: OnlyFans-style monetization for the AI version of yourself, where the creator stays in charge of every decision.

Why now

Two converging forces:

  1. Unauthorized deepfake training of adult creators is already happening at scale, in private Discord servers and gray-market clip sites. Creators have no consent, no control, no revenue, no recourse. The technology is not going back in the box.
  2. An active labor movement in the adult industry is fighting perpetual-likeness contracts at physical venues. The demand — workers want to keep ownership of their image and revoke licenses if relationships go bad — translates directly into platform infrastructure.

Likeness is the same demand turned into infrastructure: if a creator's likeness is going to be used to generate AI content regardless, the creator should be the one licensing it, setting rules, approving outputs, and earning the money.

Defensibility frame

The version that gets sued, deplatformed, and morally torched is "deepfakes for everyone." That is not the company. The defensible version is a consent-first licensing platform for verified adult creators where:

  • Every face on the platform is identity-verified (no third-party reference uploads, ever)
  • Models never leave the platform (architecture commitment, not policy)
  • Every prompt parses against the creator's license object before any model loads
  • Every output carries watermark + perceptual hash + signed metadata
  • Creators can revoke at any time

Founder

  • Role: Founding CTO / Product (founder's seat). Direct CTO experience.
  • Domain proximity: Outside-in technical founder. Not personally an adult industry insider, but has a small network of insider friends who can potentially make introductions. [Yellow Flag] — heavy reliance on the Creator Relationships & Operations cofounder hire to bridge this gap. Founder-market fit on the cultural axis is moderate, not strong.
  • Time commitment: Full-time on close of pre-seed.
  • Existing work: Pre-seed planning artifacts complete. Mockup is built and shareable. No code in production.
  • Decision not to lead fundraising: Deliberate. Recruiting a fundraising-experienced CEO is the first cofounder priority.

Market

  • Customer (creator side): Verified adult creators currently monetizing on OnlyFans, Fansly, JustFor.fans, Manyvids, and direct channels. Initial cohort is 5–10 invite-only creators willing to work closely on the concierge prototype.
  • Customer (fan side): Existing subscribers of those creators. Likeness is not a discovery platform; the creator brings the audience.
  • Geography: US-first launch, geofenced. UK/EU expansion later, jurisdiction by jurisdiction.
  • Discovery state: Pre-discovery. No structured creator interviews yet. Reviewer checklist (docs/reviewer-checklist.md) and mockup are built, but not yet in front of creators. This is the largest validation gap.

Business model

  • Subscriptions, tips, PPV, custom requests (industry-standard ~20% platform take)
  • AI generation as compute credits with creator-set markup (the new revenue line)
  • Submission fees for fan-submitted favorites to the creator's approval queue
  • Premium creator tools (analytics, takedown monitoring, advanced approval workflows) as paid add-ons or revenue share

Illustrative subscriber tiers: $15 (real content only) → $25 (+AI gallery) → $50 (+AI generation credits) → $100 (premium with submission privileges) → $200 (self-insert tier, post-MVP).

What's in MVP (concierge phase)

  • 5–10 invite-only creators
  • Manual model training (per-creator LoRA on a frozen open-weights base, working assumption Flux.1 Dev or SDXL with a defensible community fine-tune)
  • Still images only
  • Subscriptions + compute credits via adult-friendly payment processor
  • License engine + approval queue + watermarking + manual moderation

Explicitly NOT in MVP

Video generation, voice cloning, public discovery/search, open API, fan self-insert, mobile app distribution, unrestricted creator collaborations.

Capital plan

  • Pre-seed target: $1.5M for 18 months
  • Six-seat founding team (CTO/Product = founder, CEO, Compliance & Legal Lead, ML Lead, T&S Lead, Creator Ops Lead)
  • Personnel-dominated budget (~63%)
  • ~11% reserve
  • Adult-friendly payment processors are the single most likely cause of needing a bridge round before seed

What's already non-negotiable

These have been settled and won't be relitigated:

  • No model export, ever (architecture-level)
  • No third-party reference uploads
  • License-gated inference
  • Per-output provenance
  • Per-creator adapter isolation
  • Hard-block: minors, age-ambiguous content, public figures, nonconsent scenarios, "leaked tape" framing
  • Conservative content rules (categories like nonconsent fantasy, incest roleplay, "barely legal" are off the table — adult-friendly processors require this posture)

Hard questions named honestly

  • Revocation can't recover downloaded files. Pitch is "more control than you have today, plus a serious enforcement toolkit," not "perfect rights management."
  • Payment processors are a single point of failure. Stripe and PayPal will not work. Need adult-friendly high-risk processors with redundancy.
  • Creator trust is the biggest non-legal risk. If creators feel the platform replaces them with AI versions of themselves, the product fails.
  • Reserve can still get eaten. A single severe processor or legal incident can burn most of the $170K reserve in one quarter.

Founder Q&A (this session)

Question Answer
Domain proximity to adult industry Outside-in technical founder with small network of insider friends who can introduce
Creator conversations to date None yet — pre-discovery
Research scope priorities Adult AI likeness platforms, consent/likeness rights infrastructure, adult payment processor landscape (deprioritize: mainstream AI companion platforms and OnlyFans-substitute frame)

Flags

Red Flags: - Zero customer discovery to date. The platform's most load-bearing assumption — that adult creators want this product enough to participate at the rules described — has not been tested with a single structured creator conversation. Validation phase will weight creator interviews heavily.

Yellow Flags: - Founder-market fit is moderate, not strong. Outside-in technical founder is appropriate for the platform/architecture work, but creator trust will rest on the Creator Ops cofounder hire and the network they bring. The company's cultural fit is contingent on a hire that hasn't happened. - Compliance & Legal Lead and Founding CEO are budget-fragile. Both have specialty profiles where market rate is roughly double the cofounder-modest cash budgeted. Recruiting risk is real. - Payment processor risk is named but not yet validated. No processor relationship is in place; no processor due-diligence pre-conversation has happened. This is the single most likely cause of a bridge round before seed.

Sources

  • /Users/joemattie/Desktop/Likeness/likeness/docs/founder-brief.md — primary source for product, business model, risk framing, and capital plan
  • /Users/joemattie/Desktop/Likeness/likeness/docs/budget.md — capital plan and burn modeling
  • /Users/joemattie/Desktop/Likeness/likeness/docs/cofounder-jds.md — six-seat team structure
  • /Users/joemattie/Desktop/Likeness/likeness/docs/ml-lead-technical-brief.md — technical posture
  • Founder Q&A, this session (2026-05-09) — domain proximity, discovery state, research scope