⏻ENTITY.ID OS v1.0 · INVESTOR SESSIONproc: title · --:--
entity-os — boot● READY
ENTITY.ID · JULY 2026
The venture operating system.
Where incorporation is a feature — not the front door. Form in minutes. Operate immediately. Incorporate when you need to.
$for founders, collectives, and AI agents
Open cold, no "hi we're". Land the category slowly: Entity.ID is the venture operating system. Then the thesis in one line: incorporation is a feature, not the front door. Everything on the following slides is that sentence, proven.
why-now.log — three curves crossing● TAILING
Three curves crossed in 2025–26. Nobody built for the intersection.
CURVE_1 · CREATIONCENSUS BFS
0M
US business applications in 2025 — all-time record, accelerating into 2026
CURVE_2 · AI FOUNDERSSTRIPE / INC.
+0%
Stripe Atlas incorporations YoY, Q1 2026 — first $1B one-employee company predicted for 2026
CURVE_3 · AGENT MONEYFORBES / X402
~$0M/yr
moved by ~69K agents on one protocol; every major payment network shipped agent rails in 18 months
$agents can pay. they still can't legally exist.
Last year Americans filed 5.6 million new business applications — a record. AI is minting founders: Atlas up 130% year over year; Anthropic's CEO predicts the first billion-dollar one-employee company this year. And in eighteen months every major payment network — Visa, Mastercard, Stripe, Google — shipped rails for AI agents to pay. Seventy thousand agents already move six hundred million dollars a year. Beat. Not one of them legally exists.
stack-trace — where the venture stack begins● FAULT
The entire venture stack starts after the moment that matters.
THE INCUMBENT STACKIDEA → WEEKS OF PAPERWORK → THEN TOOLS
FILING GATE
CARTA · ANGELLIST · BANKING — LOCKED
ENTITY.IDOPERATIONAL FROM THE IDEA
treasury · fundraising · governance · compliance — live from minute one; incorporation = optional module
ERR_1Formation platforms stop at paperwork — LegalZoom: $756M/yr of pure filings in a ~4%-CAGR market (SEC)ENDS AT PDF
ERR_2Ops tools can't start early — Carta requires a certificate of incorporation to onboard; AngelList funds only incorporated companiesGATED
ERR_3So the riskiest phase runs on handshakes — and courts call that a general partnership with personal liability(Lido, 2025)EXPOSED
Carta won't onboard you without a certificate of incorporation. AngelList can't fund you without one. And the platforms that produce the certificate — Atlas, LegalZoom, Clerky — end at the filing. So the riskiest months of every venture — money changing hands, no roles written down — run on handshakes and someone's personal bank account. Courts have started treating those groups as general partnerships. Personal liability. Pause, let it sit. Point at the diagram: their bar stops at the filing gate; ours starts at the idea.
kernel — the primitive● RESOLVED
A venture is a disclosure + a registry — not a filing.
What makes an entity real isn't the state's PDF. It's a verifiable public record of who it is, who's behind it, what rules it runs by, and where the money sits. Make the record the primitive, and:
>operations start at the record, not the filing — day-zero ops
>incorporation compiles into a view of the record, generated on demand — a feature, not a gate
>anything that can carry a record can be a venture — including an AI agent(KYA)
PROOFRegistries are products: Estonia e-Residency — 132K+ e-residents, €0M/yr to the state (ERR News)PRECEDENT
The insight: a venture is a disclosure plus a registry, not a filing. Once the record is the primitive, three things follow — operations start at the record; incorporation becomes a view you generate when a counterparty demands it; and anything that can carry a record can be a venture, including an AI agent. Estonia already proved a registry is a product: a hundred twenty-five million euros a year.
boot — tarka-labs.public.entity.id● BOOTING…
Boot a venture in minutes. It runs itself from minute one.
SIMULATED DEMO · TARKA LABS IS A DEMO VENTURE, NOT A CUSTOMER · CLICK TO PLAY
◇Formation7 steps: name → members → ownership → constitution → register…
▣Incorporationinstall when a counterparty demands it — record & history carry overOPTIONAL
Shift gear — demo energy. Seven guided steps — name, members, ownership split, constitution — and you have a real venture on a tamper-proof public registry at tarka-labs dot entity dot i-d. The moment it's registered, it runs: treasury live, fundraising live, governance live, compliance live. An AI back-office handles the filings and the minutes. Incorporation? It's a module. You install it when a counterparty demands it. A feature — not the front door.
kya — agents with standing● FIRST-CLASS
The first venture layer where agents are first-class.
SIDE_AAI runs the venture's back office — ops cost approaches zero; a team of one operates like fiftySHIPPING
SIDE_BAgents join ventures as members / operators — verifiable standing, anchored to accountable humans: KYA, Know Your AgentTHE WEDGE
GAPIdentity is standardizing (A2A, W3C, IETF) · payments are a land-grab (Visa vs Mastercard vs Coinbase) · legal standing: emptyUNSERVED
ALTClosest offer (MIDAO): $9,500 + $2–5K/yr, weeks of concierge, mandatory human kill switch — the agent is never the principalBROKEN
REGEU AI Act Art. 50 agent-disclosure applies Aug 2026 · FINRA's first AI-agent risk section · NIST RFI on agent identity — compliance-by-registration is what they're describingTAILWIND
Now the part that makes this venture-scale. Slow down — thesis voice. Agents got payment rails; the first AI to make a million dollars has no lawful way to own it. On Entity.ID, an agent registers as — or within — a venture, anchored to accountable humans, with verifiable identity and a machine-readable compliance model. KYA — Know Your Agent — KYC for the agent economy. Regulators are already converging here: EU disclosure rules bite in August; FINRA just shipped its first AI-agent guidance.
market.calc — tam expansion● COMPUTED
A paperwork market becomes an operating-system market.
Carta, on equity ops alone — where operations revenue already lives
NEW MARKETMARKETSANDMARKETS
$0B
AI agents by 2030 (46% CAGR); $15T of B2B spend agent-intermediated by 2028 (Gartner)
VOLUMECENSUS / UK CH
0M
US nonemployer businesses + 5.6M applications/yr — the underserved 80%
Three concentric markets. The beachhead is formation spend today — twenty-five billion going to forty-seven. The expansion is the operations layer, where Carta built four hundred forty million of ARR after the filing — we move it to day zero. And the new market is the agent economy: fifty-plus billion by 2030, forty-six percent CAGR. Underneath it all, volume: five and a half million applications a year and thirty million nonemployer businesses the incumbents never see.
telemetry — live registry metrics● AWAITING INPUT
The registry publishes its own numbers.
[FILL] ventures registered[FILL] MAU / WAU[FILL] treasury volume under management[FILL] agents registered[FILL] waitlist / partners
The e-Residency playbook: the registry's own dashboard is the traction slide. Real numbers only — Tarka Labs is demo material, never a customer.
Founder fills before the meeting. Frame it as telemetry: a registry that publishes its own statistics, the way Estonia's e-Residency dashboard does. Lead with the strongest number; if a number is small, pair it with its growth rate.
pricing.d — protocol economics● METERED
Free to exist. Paid to operate. Priced to the record.
TIER_0Formation free or near-free— kills the $500 gate; maximizes records on the registryGROWTH LOOP
SAASLive modules— treasury, governance, compliance, fundraising tooling: the Carta-shaped revenue, moved to day zeroRECURRING
ANCHORIncumbents charge $297–$9,500 to create entities and $2,800–$75K/yr to operate themFIELD DATA
OPTIONAL SKETCH: IF THE REGISTRY OPENS TO THIRD-PARTY REGISTRARS / VERIFIERS, USAGE-METERED PROTOCOL FEES ON REGISTRATIONS AND STANDING CHECKS ACCRUE TO OPERATORS — AN OPTION, NOT THE PLAN.
Business model: free to exist, paid to operate. Formation is free — that's the growth loop; every record makes the registry more valuable. SaaS on the live modules is the Carta-shaped revenue moved to day zero. And per-event: incorporation, filings, and KYA verification fees every time anyone checks a record. If asked about protocol economics: usage-metered protocol fees are an option if the registry opens to third-party registrars — an option, not the plan. Avoid crypto vocabulary.
releases — land at day zero, expand along the life● SCHEDULED
Land at day zero. Expand along the venture's life.
NOWFormation + live modules + public registry— shippedv1 · LIVE
NEXTIncorporation module (jurisdiction partners) · banking / treasury rails depth · KYA verification API for counterpartiesv2
THENAgent-run ventures at scale — compliance-by-registration as the default trust check of agent commerce · third-party apps on the protocolv3
[FILL] dates & sequencing from the founders' actual plan
Three releases. Now — shipped: formation, live modules, the public registry. Next: the incorporation module with jurisdiction partners, deeper treasury rails, and the KYA verification API so counterparties can check standing programmatically. Then: agent-run ventures at scale, where compliance-by-registration becomes the default trust check of agent commerce.
If any credential touches registries, identity, or corporate services — it leads.
Founder fills. Keep to names, one-line credentials, one proof point each. Anything touching registries, identity, or corporate services leads the slide.
install — the round● AWAITING INPUT
[FILL] $X at Y — round & terms
$use of funds: jurisdiction / incorporation partnerships · KYA verification API · registry growth loops (free formation)
Every venture tool starts after incorporation. We start before — which means we start everything.
$entity.id — the venture operating system
State the round and the three uses of funds. Then close — slow, eye contact: every venture tool starts after incorporation. We start before it. Which means we start everything. Beat. Entity.ID — the venture operating system.