Skip to content

Dapper Labs

A venture studio at the intersection of emerging technology and culture. We build consumer products on Flow — the network we created — with a portfolio spanning digital collectibles, consumer finance, AI agents, and virtual entertainment.

~$22M Annual Revenue
975M+ Transactions on Flow
40M+ On-Chain Accounts
~90 Team Members

Start With These

  1. Company Overview — The thesis, the three bets, why we exist
  2. Company History — Axiom Zen to CryptoKitties to Flow to today
  3. Culture and Values — Vision, mission, "why Web3"
  1. Team Structure — Who does what, who to ask
  2. How We Work — Art of the Bet, MSCW, ICE, Bold Beats
  3. AI at Dapper — The intelligence architecture
  4. Loop Herding — How we build autonomous systems
  5. AI Cutting Edge — What AI agents can do now

What We Build

Dapper operates as a venture studio with five organizational units: Established (revenue-generating), Big Bet (proving their market), Sandbox (testing hypotheses), and Protocol (the substrate). See Venture Studio Thesis for the full framework.

Established Business Units

LIVE — $10.8M/yr

NBA Top Shot

The flagship. Licensed NBA digital collectible moments. Gateway product for the portfolio — 86K single-product users, 108 XL whales generating 66-83% of revenue.

Now: Playoffs Sprint · Strategy · IPFS Permanence

HIATUS — $6.3M/yr

NFL ALL DAY

Licensed NFL video collectibles. 5-8x seasonal revenue swing. Currently in NFLPA contract renegotiation — product on hold pending resolution.

Now: Renegotiation Status · NFL Partnership

LIVE — $3.6M/yr

Disney Pinnacle

Licensed Disney character pins. Independent audience (58.5% no sports overlap). Trading as the engagement engine — Genesis Keys just launched.

Now: Trading Launch · Product Briefs · Strategy Inputs

Collectibles Portfolio

Strategic Plan · Fee Structure · Asia Expansion · Data Insights

Big Bet

BIG BET — 21 PEOPLE

Consumer Finance

Peak Money, FCM, Flow Vaults. Three decoupled layers: app on EVM yield, protocol, future gamified finance. Led by Bart Bobnis & Alex Hentschel.

Now: Roadmap · Consumer Finance Overview

Sandbox

SANDBOX — 2 PEOPLE

Octopus Rodeo

AI-native products. CryptoKitties (~300K players), Miquela ($550-600K rev). Led by Alan Carr & Renan Sgorlom.

SANDBOX — 2 PEOPLE

Riptide

Autonomous on-chain AI. Agent infrastructure research. Led by Jan Bernatik & Navid TehraniFar. Layer 4 coordination thesis.

Now: Roadmap · Flow Protocol

Protocol

PROTOCOL — 11 PEOPLE

Flow Core

The blockchain agents choose. AI-feature ready. Merged into Dapper from Flow Foundation. Led by Dieter Shirley & Kan Zhang.

Now: Flow Basics · Exchanges


Data-Driven Decisions

Every product decision should reference a validated finding. 22 findings from BigQuery analysis, each cross-validated across multiple waves.

Section The Question It Answers
:material-funnel: Pipeline & Funnel Health Where do users drop off between signup and whale?
Whale Economics Who spends, how concentrated is revenue, why do whales leave?
Engagement Levers What actually retains users — and what doesn't?
Product Operations When does revenue peak, how should drops be timed?

The 10 Governing Numbers

The Collectibles Overview opens with the 10 numbers that define the business. Start there.


Playbooks & Process

Step-by-step guides for running the business. If someone would DM a colleague to ask "how does X work?" — the answer should be here.

Decision Log — major product decisions with date, who decided, why, and what changed.


Technical Foundation

System What It Is Deep Dive
Atlas Next-gen platform — campaign builder, feature flags, AI-driven ops Atlas Platform
Flow Layer-1 network — Cadence + EVM, Forte, 975M+ transactions Flow Basics
FCM Zero-liquidation lending — automated rebalancing via Forte FCM Technical
Peak Money Backend Payment stack — Stytch, Sardine, HIFI, Temporal, Crossmint Architecture
AI & Infra SRE/DevOps evolution: campaign builder, Heimdall, KAAOS daemon, Morning Brief AI & Infrastructure
Geppetto Autonomous product intelligence loop Architecture
Data BigQuery semantic layer, production_sem_open.* Data Infrastructure

How This Wiki Works

Agent-maintained, human-audited. AI agents ingest signals from meetings, Slack, GitHub, Linear, and Drive — then update wiki pages with provenance tags. A red team agent audits every page for confidential information before deployment.

Tier Meaning
CANONICAL Verifiable system-of-record fact
DECISION Human authority call
DERIVED Agent synthesis from multiple sources
HEARSAY Someone said something

Can't find something? Flag the gap in #team-wiki or add it yourself. The wiki's completeness determines the AI coach's usefulness.