Company profile · 3 min read
TwentyTwo: Reviewable screening for the AI-enabled bioeconomy
TwentyTwo builds biosecurity infrastructure, starting with AI-powered customer screening and previewing sequence and model safeguards.
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
Published · Updated
TwentyTwo is building a defensive layer for the bioeconomy, starting with AI-powered screening. It fits life-science companies and AI-biology builders that need to make dual-use review part of their operating workflow rather than a last-minute manual check.
What it does
TwentyTwo’s current homepage describes three product areas. Automated Customer Screening is marked available and combines AI agents with traditional data sources to gather background intelligence, while the customer makes the decision and receives an audit trail. Functional Screening, which is intended to identify biological sequences that evade BLAST-based screening, is marked preview. Model Safeguards for AI-biology developers is also marked preview.
That product ordering matters. The current buyer can evaluate a screening workflow today; the more ambitious sequence-function and model-safeguard layers are not presented as generally available. The homepage frames the timing around cheaper synthetic biology, cloud labs and stronger AI models, and says new regulation is creating a need for defensible screening. Those are the company’s why-now claims, not a legal opinion about a buyer’s obligations.
The current biosecurity record also sits next to an older YC launch. The directory page preserves Mohi, a debugging assistant for AI agents that visualized traces and diagnosed failed steps. That is useful company history, but the current homepage has a clear biosecurity identity and should be the basis for a current buyer decision.
Evidence and founder context
TwentyTwo’s homepage says the team built biosurveillance across 27 U.S. sites monitoring more than 13M people, contributed AI biosecurity evaluations adopted by frontier AI labs and published work on viral-evolution prediction. These are company-published team and research claims. The current YC profile identifies Evan Seeyave as founder and CEO but provides little additional professional background.
Pricing is not public. A buyer should ask which sequence libraries and data sources are covered, how human review is recorded, what evidence is retained for an audit and how screening decisions are escalated when the model is uncertain.
Editorial take
I would shortlist TwentyTwo for a DNA-synthesis, biotech or AI-biology team that needs a reviewable screening process and can start with the available customer-screening module. I would not buy it on the promise of the preview modules. The useful decision is whether the audit trail and background research improve the team’s current screening bottleneck.
Quick facts
| Field | Sourced detail |
|---|---|
| Product | AI infrastructure for biosecurity, starting with customer screening |
| Buyers | Life-science companies, DNA providers and AI-biology builders |
| Current availability | Customer screening marked available; functional screening and model safeguards marked preview |
| Company evidence | Biosurveillance and research claims published on the homepage |
| Pricing | Not publicly listed |
Sources checked
| Source | Checked |
|---|---|
| YC profile | 2026-09-19 |
| TwentyTwo homepage | 2026-09-19 |
| Mohi launch record | 2026-09-19 |
Cohort context
TwentyTwo is listed in Summer 2025. In our 2026-09-18 directory snapshot, 11 of 166 listed companies in that cohort have YC’s primary industry label Healthcare (6.6%). This is a current-directory comparison, not an original intake count or a performance ranking. Nine-cohort dataset.
Public website snapshot
Observed 2026-09-19T16:18:27.547Z in raw homepage HTML. This records visible metadata and advertised links, not agent execution or product quality.
| Signal | Homepage observation |
|---|---|
| Product description metadata | Observed |
| Canonical link | Not observed in this response |
| H1 or H2 heading | Observed |
| Typed structured data | Not observed in this response |
| Docs/developer link | Not observed in this response |
| Pricing link | Not observed in this response |
| llms.txt link | Not observed in this response |
| Markdown alternate | Not observed in this response |
Public observations · Collection method. Missing links here do not establish that a capability or file is absent elsewhere.
About the author
Cofounder of Lazyweb (a16z Speedrun). Previously grew subscriptions and built native ads at Duolingo. Writes First1000 (90,000+ subscribers).
I cofound Lazyweb and publish Mudpie. This is an owner-written publication, not an independent testing organization. Research notes distinguish observations, sourced reporting and editorial judgment.
First1000 ↗ · X ↗