Reef · reef.bio
← reef.bio
For investors

Cursor for organisms.

In 1981, AutoCAD put engineering design in the hands of individual engineers. In 2022, Cursor put AI-native coding in the hands of every developer. Reef is building the same shift for biology: describe the behavior you want, like "bacteria that sense glucose above 5mM and produce insulin for six hours," and get a simulated, validated, build-ready genetic design. The AI handles the expertise. You review and ship biology.

The market

A proven market, still engineering organisms with punch-card tools.

Synthetic biology is a $19B market growing to $112B by 2033. Every company in it engineers organisms. None of them have a Cursor.

$7B → $15B
Synbio tools and software (the direct market) by 2033, ~19% CAGR.
Source: DataInsightsMarket
$19B → $112B
Total synthetic biology by 2033, 17 to 23% CAGR.
Source: Grand View · Coherent
$200M ARR
Benchling proves the spend: 200k+ scientists pay for biology software, none of it AI-native circuit design.
Benchling, ~$6.1B valuation (2021)

Market sizes are third-party estimates, shown for context. Detailed TAM/SAM/SOM modeling and our revenue projections are in the deck.

Why now

Three things became true in the last 18 months.

01

Models got good enough

Foundation models for biology (Evo2, ESM3, AlphaFold3) made natural-language intent into structured circuit design tractable. Eighteen months ago it wasn't.

02

Living medicine got approved

The first FDA-approved live biotherapeutics created a category. Every company racing into it needs circuit-design tooling.

03

The gap stayed open

Incumbents went wider (more of the lab stack), not deeper. The circuit layer, the wiring above proteins, remains unoccupied.

The problem

Expertise is the bottleneck. Logistics is the tax.

The expert bottleneck

~50,000 synthetic biologists. Few can design circuits.

Most know what they want to build but lack the systems-biology expertise to design it. The right AI layer lets any trained biologist design at a computational-PhD level.

The design-build disconnect

40 to 60% of iteration is logistics, not science.

Design lives in a notebook and never connects to synthesis, sequencing, or characterization. A two-week validation takes six.

The data silo

The same part is re-measured by 1,000 labs.

Characterization data is trapped in individual notebooks and never aggregated, so everyone simulates from decade-old, mismatched parameters.

Why it's defensible

The moat compounds with every design.

The data flywheelprimary moat

Every circuit designed, built, and characterized in Reef returns a validated data point that sharpens the next prediction. Over years, this becomes the largest database of real circuit behavior, which is irreproducible without running millions of experiments.

The standardformat moat

Built on SBOL, the open synbio standard. Interoperable with the whole ecosystem; proprietary formats are dead in this community.

The communitydistribution

An open design canvas makes Reef's format the default, with the iGEM to professional pipeline as a recurring, low-cost acquisition channel.

Structural lock-inspositioning

The natural incumbents are structurally barred from this layer: selling researcher design tools would cannibalize their own foundry or licensing revenue. (Detail in the deck.)

Business model

Software margins, plus a cut of the build.

Two stacking streams at ~90% gross margin, with an optional co-development tier later.

SaaS subscriptions

Tiered plans from a free entry tier through team and enterprise, each with a monthly credit bundle for AI generation and simulation. (Exact tiers and credit pricing in the deck.)

Transaction take rate

A percentage of every synthesis and sequencing order routed through the integrated provider network, additive to subscription revenue on every tier.

Where we are

Live engine today. A staged path to the platform.

The behavior engine and circuit validator run today; you can try them on reef.bio. The roadmap closes the design-build-test loop, then expands up the abstraction stack.

Today · V0

Circuit validator + live behavior engine

Upload a design, get failure prediction before synthesis. Running in the browser now.

2026 · V1 · Circuit by Reef

The AI-native circuit IDE

SBOL visual editor, AI copilot, simulation, registry search, one-click synthesis ordering.

2027 · V2-V3

Repository + metabolic pathways

The design-build-test loop closes; versioning, collaboration, and pathway design.

2028+ · V4-V5

Organism design + open platform

Whole-organism design, the living-medicine module, and an open API on the data layer.

The team

A biologist and an AI engineer: the exact pair this needs.

Domain · product · regulatory

A clinician-scientist who knows when an AI's biological output is nonsensical, owns regulatory strategy (FDA / IND / LBP), and opens academic and clinical relationships.

Technical · AI engineering

An AI engineer building the copilot, the simulation backend, the visual editor, and the provider integrations.

Building the Cursor for organisms takes someone who thinks like a biologist and someone who builds AI systems. Neither alone would build the right product. Full founder bios are in the deck.

The deck

Want the full picture?

Request the investor deck: the detailed market model, financials, competitive teardown, moat, and the round. We'll send a tracked link and follow up.

  • Market model + revenue projections
  • Full competitive analysis & defensibility
  • Roadmap, team, and the raise

We'll only use this to send the deck and follow up.