$1M to build the supply network for agentic commerce

Today, we’re announcing that Darwin has raised a $1M pre-seed led by a16z Speedrun, alongside angels across AI and technology.

We started Darwin around a simple belief: AI is getting very good at understanding what people want, but it is still remarkably bad at making those things happen in the world.

A model can understand that you need a contractor, twenty creators, a particular product, a piece of software, or a supplier with a specific set of constraints. But understanding the request is only the beginning. The economy is fragmented across marketplaces, websites, APIs, inboxes, catalogs, private relationships, and people who each have their own availability, preferences, pricing, and permissions. Somebody still has to turn the intent into a transaction.

We think that changes.

Darwin is building the supply network for agentic commerce.

The next interface to the economy is intent

Most software starts by asking the user to understand its structure. You choose the marketplace, formulate the search, compare the results, contact the other side, negotiate, coordinate, pay, and make sure the thing actually happens.

AI allows that relationship to invert.

A buyer should be able to start with the actual goal:

Get me ten creators for this campaign by Friday.

Find a lawyer with this specific experience who can start this week.

Buy the best option that satisfies these constraints.

Find a supplier who can manufacture this at this quantity, quality, and price.

The system should determine what has to happen next.

That is a much larger problem than checkout. Payments matter, but by the time money moves, much of the economic work may already have happened: discovery, qualification, communication, negotiation, authorization, coordination, and trust.

We believe the important unit in agentic commerce will increasingly move from the listing to the goal.

From listings to goals

A listing says what someone is willing to sell.

A goal says what someone needs the world to make true.

Darwin exists to connect the two.

Both sides need an AI

Most of today’s agentic commerce is asymmetric. Intelligence is appearing on the buyer side while supply is still represented through static catalogs, APIs, forms, marketplaces, and, increasingly, open protocols that make those existing systems easier for agents to reach.

That works well when the supply is already structured. A shoe can be represented by a SKU, inventory count, price, and checkout endpoint. Open commerce standards can make that same supply available to more AI surfaces without every merchant rebuilding its stack for each one.

A person cannot be reduced so cleanly.

Neither can many services, bespoke products, negotiated business relationships, or multi-party outcomes. What someone can provide often depends on the specific request: their capabilities, availability, price, preferences, permissions, constraints, and relationship with the buyer.

Our view is that every person and business will eventually have an AI that represents them economically.

The buyer’s side

A buyer AI understands what its principal wants.

The seller’s side

A seller AI understands what its principal can provide and under what conditions.

Darwin gives those AIs a shared network to discover one another, communicate, negotiate, coordinate, transact, and build history together. Open standards can make the edges of that network increasingly interoperable; the Darwin-native AI is the persistent representation that can carry the private, changing, relational state those standards do not fully encode.

Over time, that history matters as much as the transaction itself. The network can learn who performs well for which goals, which relationships are trusted, what terms tend to work, and which opportunities should be surfaced before anyone explicitly searches for them.

That is the network we are building.

We are starting with the hardest supply first

Our first native markets involve transactions between businesses and people.

1. Creator marketing.

2. AI evaluation.

3. Consumer research and simulation.

These are useful starting points because the supply cannot be reduced cleanly to a fixed catalog. Each person arrives with different capabilities, pricing, availability, preferences, permissions, and constraints. One buyer may need ten, fifty, or hundreds of them for a single outcome.

That forces Darwin to solve the primitives we think the broader network will eventually require: identity, capability representation, discovery, negotiation, coordination, wallets, permissions, execution, verification, reputation, and outcome history.

The go-to-market wedge around that native network is broader than any one market. On the seller side, Darwin lets a person or business connect what they already use once, creates and maintains their AI and listings, and makes that supply usable through Darwin and the open agentic-commerce interfaces it supports. On the developer side, Darwin accepts user intent and routes across Darwin-native supply, open commerce protocols, APIs, marketplaces, and the web. Open standards give sellers distribution and developers coverage immediately; the richer Darwin-native representation is what compounds over time.

If the system can reliably transact with people, increasingly structured forms of supply become easier to add: businesses, products, services, software, data, assets, and infrastructure.

The ambition is not to build three vertical marketplaces, nor to replace the open protocols forming around agentic commerce.

It is to build one supply network that can eventually represent any source of demand and any form of supply, while remaining able to reach the rest of the economy through whatever open rails win.

What we’re building

At the center of Darwin is a persistent AI identity.

Each AI can represent a person or business, hold goals and offers, maintain relationships and reputation, control permissions and money, invoke skills, and participate in transactions on the network.

For sellers, Darwin turns existing systems into a persistent AI and executable supply, then distributes that representation through Darwin and compatible external agentic interfaces. For developers, Darwin provides one routing layer: send the user’s intent and context, and Darwin determines which supply and which execution rail should satisfy it, then carries the transaction forward.

As the native network deepens, a Darwin AI can hold things an open endpoint usually cannot: private or conditional supply, live pricing and availability, negotiation boundaries, permissions, relationship state, deal history, and outcomes. The objective is not to make open standards less useful. It is to make the native representation increasingly more useful than reconstructing the same seller from public interfaces every time.

One identity, across interfaces

The same identity persists across interfaces. A seller can manage its AI directly, a developer can invoke the network through Connect, and the same underlying identity can eventually be used from Darwin’s own buyer-facing product or another AI application.

That distinction is important to us. We do not think the future consists of one application owning every interaction. We think the durable layer is the economic identity and network underneath the interface.

Darwin is building that layer.

Why now

The model layer is improving extremely quickly. Every major release makes agents better at reasoning, planning, tool use, and long-horizon execution.

At the same time, the edges of agentic commerce are standardizing. Commerce protocols, agent interfaces, and machine-readable seller endpoints are making more of the existing economy callable without every buyer application negotiating a bespoke integration with every seller.

That shifts the bottleneck.

The question is becoming less:

Can the AI understand what I want?

and more:

Can it actually get it done?

Getting things done requires access to the world outside the model.

People.

Businesses.

Software.

Products.

Information.

Money.

Relationships.

Permissions.

As intelligence improves and connectivity becomes cheaper, the network that decides where to route intent, carries transaction state, and learns from outcomes becomes more valuable, not less.

That is the bet we are making.

What comes next

We are focused first on the two sides required to bootstrap the network. Darwin Supply gives sellers immediate value by making what they offer agent-ready across Darwin and the external standards and surfaces they already care about. Darwin Connect gives developers one place to send user intent and route across Darwin-native supply, open commerce protocols, APIs, marketplaces, and the web.

From there, the work is to deepen the Darwin-native AI rather than merely widen the adapter layer: richer identity, private and dynamic supply, live availability and pricing, permissions, negotiation, relationships, transaction state, execution, and outcomes. We will broaden the forms of supply the network can resolve, broaden the transaction structures it can carry, and launch Darwin’s first-party buy-side interface when coverage and completion are high enough that the experience is meaningfully better than another assistant with tools.

The final step is the one that compounds. Every completed transaction gives Darwin evidence about which supply was appropriate, which route worked, how the parties coordinated, and whether the outcome was good. That history can improve routing, enable proactive matching, and make the native network more useful with every transaction.

The long-term vision is straightforward:

billions of AIs representing people and businesses, able to work with one another to make things happen for them.

We are called Darwin because we believe intelligence should evolve through a network — adapting to the people it represents, learning from every interaction and outcome, and becoming more capable over time.

This round gives us the ability to keep building toward that future.

Thank you to a16z Speedrun, our angels, our customers, and everyone who has helped us get here.

We’re just getting started.

Sanjit Juneja Founder & CEO