b-12

What do they actually do

b-12 builds an AI and automation platform that turns a chemist’s plain‑language goal into a detailed experimental protocol and machine‑readable robot instructions. It’s being deployed through pharma pilots today rather than a broad SaaS rollout. The system is positioned as vendor‑agnostic and can also output human‑readable protocols for manual labs that don’t have robots YC launch, website.

A typical workflow is: a scientist describes the goal in a conversational interface; b‑12 generates a protocol and translates it to hardware commands; the user reviews and runs on existing lab automation or executes manually; run data returns to the platform to inform follow‑up experiments. Public materials emphasize pilots, demos and enterprise security (trust center) as they work toward a repeatable enterprise product YC company, website.

Who are their target customer(s)

• Medicinal chemist at a pharma R&D team: Turning an idea into a safe, detailed protocol and waiting for results slows iteration and creates chances for error. They want faster protocol generation and fewer manual cycles.
• Lab automation engineer / automation ops at a pharma or CRO lab: They translate chemists’ intents into vendor‑specific robot code and maintain brittle integrations across instruments. They want machine‑ready instructions and a single orchestration layer.
• Small biotech or academic discovery group with limited automation expertise: They lack staff and tooling to run many experiments quickly. They need human‑readable protocols and simple robot instructions to lower the barrier to executing experiments.
• Assay lead or data scientist responsible for experiment optimization: Results sit in spreadsheets and rarely flow back into planning, making optimization loops slow. They need runs and data to feed back into next‑step experiment selection.
• Compliance/QA officer at a regulated lab: They need traceability, audit trails and validation that AI‑generated protocols meet safety and regulatory requirements. They look for enterprise security materials and controls.

How would they acquire their first 10, 50, and 100 customers

• First 10: Run tightly scoped pilots with pharma/CRO teams the founders can directly support, provide hands‑on integrations and protocol validation, and align on a single measurable success metric to secure references YC launch, website.
• First 50: Turn pilots into a standard “pilot kit” (SLA, security/compliance docs, integration templates) and scale with a small enterprise sales motion targeting mid‑size pharma, CROs and well‑funded biotechs; co‑sell with automation vendors/integrators to tap their installed bases YC company, website.
• First 100: Productize onboarding and subscription pricing with strong customer success on compliance/ROI; expand channels (CROs, LIMS vendors, automation resellers) and publish validation and security artifacts to shorten procurement in regulated labs website.

What is the rough total addressable market

Top-down context:

The relevant spend sits at the overlap of lab‑automation software and AI tools for discovery. Global pharma R&D is a large top‑level pool (~$276B), while lab automation is ~$8–8.4B and AI in drug discovery is in the low‑single‑digit billions with fast growth R&D World, Grand View – Lab Automation, GMI – AI in Drug Discovery, Grand View – AI in Drug Discovery press release.

Bottom-up calculation:

Estimate 10,000–20,000 relevant buyer sites globally (pharma R&D groups, CRO facilities, scaled biotechs running or planning automation) each spending ~$200k–$500k annually on orchestration/automation software plus AI experiment‑planning tools, implying ~$2–$10B in addressable software spend. This aligns with combining the lab‑automation software share and AI‑in‑drug‑discovery software markets today.

Assumptions:

• Counts include pharma R&D sites, CRO automation facilities, and larger biotechs likely to buy orchestration/AI software.
• Average annual software spend per site spans orchestration, integrations, and AI planning/optimization.
• Hardware spend is excluded; figures represent software/platform/integration budgets only.

Who are some of their notable competitors

• Synthace: Lab‑orchestration software (Antha) that translates experimental workflows into executable protocols across instruments; overlaps on vendor‑agnostic protocol design and automation.
• Strateos: Cloud‑lab plus lab‑control software that lets teams design, schedule and run automated experiments locally or in Strateos facilities; an alternative execution and data layer.
• Emerald Cloud Lab (ECL): Remote, fully automated lab where users submit protocols and ECL executes and returns data; substitutes for in‑house automation and orchestration.
• PostEra: AI‑first medicinal chemistry platform focused on synthesis planning and design; overlaps on what/how to make, but centers on route selection rather than robot code generation.
• Opentrons: Affordable lab robots with protocol design software; appeals to small labs that may prefer simple on‑device tooling over an enterprise, multi‑vendor orchestration layer.