Drylab AI
AI platform automating biomedical research for life science teams, from data analysis to model training.
Website: https://thedrylab.com/
Cover Block
Publicly reported
| Field | Value |
|---|---|
| Name | Drylab AI |
| Tagline | AI platform automating biomedical research for life science teams, from data analysis to model training. |
| Headquarters | San Francisco [Crunchbase] |
| Founded | 2025 [Crunchbase] |
| Stage | Seed [The Business Times, October 2026] |
| Business Model | SaaS |
| Industry | Healthtech |
| Technology | AI / Machine Learning |
| Geography | Global / Remote-First |
| Growth Profile | Venture Scale |
| Founding Team | Huy Phan (Henry), Nhi Nguyen, John Phan [LinkedIn] |
| Funding Label | Seed |
| Total Disclosed | ~$3,000,000 [The Business Times, October 2026] |
Links
Publicly reported
- Website: https://thedrylab.com/
- LinkedIn: https://www.linkedin.com/company/drylab-ai
- GitHub: https://drylab-bio.github.io/
Summary and Signal
PUBLIC Drylab AI is building software that uses AI agents to automate parts of biomedical research, and it merits investor attention now because it has already secured a US$3 million seed round led by Vertex Ventures in a category where enterprise demand is tied to labor-intensive, regulated scientific workflows [The Business Times, October 2026]. The company was founded in 2025 and describes itself as operating between Singapore and the US, with San Francisco listed as headquarters in the structured research record [The Business Times, October 2026].
The product thesis is straightforward: Drylab says its platform can discover datasets, train custom models, and run bioinformatics pipelines through a web app and desktop app, with an enterprise deployment option that can run inside customer infrastructure [Drylab] [The Business Times, October 2026]. That matters because the differentiation appears to rest less on a general-purpose model layer and more on workflow automation for biopharma and biomedical teams, plus security features such as local desktop operation and encrypted data handling, although most of these product claims remain company-sourced [Drylab].
The founding team is presented publicly as John Phan, Huy Phan, and Nhi Nguyen, with John Phan identified as CEO in the funding coverage and on LinkedIn, while other role descriptions are less consistently corroborated across sources [The Business Times, October 2026] [LinkedIn, retrieved 2026]. The available public record supports who the founders are, but it is thinner on prior operating history that would let an investor underwrite enterprise go-to-market or scaled life sciences execution with confidence [LinkedIn, retrieved 2026].
Over the next 12 to 18 months, the key public markers to watch are whether Drylab converts early category interest into named enterprise adoption, whether security and compliance claims mature into verifiable certifications or referenceable deployments, and whether product breadth can be translated into a repeatable buying motion rather than a wide feature surface [The Business Times, October 2026] [Drylab].
No independent source found -- This section relies on one independent funding report and multiple company-sourced product claims, with partial founder corroboration from LinkedIn.
Taxonomy Snapshot
| Axis | Value |
|---|---|
| Stage | Seed |
| Business Model | SaaS |
| Industry / Vertical | Healthtech |
| Technology Type | AI / Machine Learning |
| Geography | Global / Remote-First |
| Growth Profile | Venture Scale |
| Founding Team | Co-Founders (3+) |
| Funding | Seed, total disclosed ~US$3,000,000 |
Company Overview
PUBLIC
Drylab AI is a newly formed life sciences software company that appears to be building around one narrow promise: reducing the manual work involved in biomedical analysis and model-building workflows. Public company materials describe a San Francisco-based business founded in 2025, with a SaaS platform focused on biomedical research automation, spanning dataset discovery, custom model training, and bioinformatics pipelines [Crunchbase] [Drylab].
The public record is still thin, which is typical at this stage. Crunchbase lists Drylab AI as founded in 2025 and headquartered in San Francisco [Crunchbase]. On its website, the company presents itself as an AI bioinformatics platform for biomedical research, offers both web and desktop applications, and says the product supports use cases including biomarker discovery and patient stratification [Drylab].
The clearest milestone visible in company-controlled sources is product availability rather than a long operating history. Drylab's website shows a live product surface and describes the platform as combining no-code workflows with code generation and pipeline execution for more advanced users [Drylab]. Crunchbase also identifies the business as seed-stage and venture-backed, which fits a company still defining category position and early go-to-market motion rather than one with a long disclosed operating track record [Crunchbase].
No independent source found -- This section relies on Crunchbase and company website materials, with several material company descriptors supported only by company-controlled sources.
The Product and the Stack
Public record plus analysis
Drylab AI is positioning the product around a fairly specific pain point: biological data work that is fragmented across datasets, tools, and specialist scripting. On its website, the company says the platform automates biomedical research workflows end to end, including dataset discovery, custom model training, and one-click bioinformatics pipelines, and that it is built for both no-code users and experienced bioinformaticians [Drylab]. The same materials describe use cases in biomarker discovery and patient stratification, alongside both a web app and a desktop app, which suggests the company is trying to serve regulated research teams that want convenience without giving up local control over sensitive data [Drylab].
The more differentiated claim is not generic workflow software but agentic operation inside research environments. The Business Times reported in October 2026 that Drylab can be deployed within the infrastructure of a pharmaceutical company or research institution to develop autonomous AI agents for scientific research [The Business Times, October 2026]. On the company side, Drylab says its desktop mode keeps data on the user machine and that its "Drylab Vault" security layer includes MFA and AES-256 encryption [Drylab]. Those claims fit the biopharma buyer the company is targeting, but they remain mostly company-described rather than independently validated in a public product review or technical audit [Drylab] [The Business Times, October 2026].
A separate technical signal is DRYLAB-Bench, a biosecurity benchmark hosted on the company's research site and described as being under review for ICLR 2027 [drylab-bio.github.io]. According to that material, the benchmark evaluates LLM-designed protein mutations against wet-lab measurements and includes 30 scenarios, 15 frontier models, and agentic tool use [drylab-bio.github.io]. That does not by itself validate commercial product performance, but it does indicate the team is trying to ground its agent claims in domain-specific evaluation rather than general-purpose model rhetoric [drylab-bio.github.io].
No independent source found -- Material product details rely primarily on company website claims, with limited independent corroboration from The Business Times.
The Market They Are Entering
PUBLIC
The market matters now because drug developers and biomedical researchers face a plain imbalance: biological data volumes are rising faster than the specialist labor available to process them, while AI tooling has become credible enough to move from assistive analysis toward more automated research workflows [The Business Times, October 2026].
The public record for Drylab AI does not include a named third-party TAM, SAM, or SOM estimate tied directly to the company’s category. That limits precision, and it matters, because the startup sits at the intersection of several adjacent markets rather than a single clean software budget line. The most supportable public framing is therefore analogical: Drylab AI appears to target software spend within biopharma R&D informatics, bioinformatics workflow tooling, and regulated enterprise AI deployment for research organizations, based on its stated product scope and customer focus [Drylab] [The Business Times, October 2026].
Demand signals in the available sources are narrow but directionally consistent. The Business Times reports that the company is aimed at biopharmaceutical and biomedical organizations and can be deployed inside the infrastructure of a pharmaceutical company or research institution, which suggests a market defined less by general-purpose AI adoption and more by institution-grade research environments with security, compliance, and workflow requirements [The Business Times, October 2026]. On the company side, Drylab describes a platform that automates dataset discovery, custom model training, and bioinformatics pipelines, with use cases in biomarker discovery and patient stratification, both categories that sit close to higher-value translational and clinical research budgets if the workflow performs as advertised [Drylab].
The tailwinds are easier to identify than the market size. Multi-omics research, model-building, and bioinformatics remain operationally fragmented, and Drylab’s public materials are explicit that the product is trying to collapse tool sprawl into one interface, citing integration with 800-plus bioinformatic tools and support for both no-code and expert users [Drylab]. If that claim holds in practice, the relevant demand driver is not only AI enthusiasm, but labor compression: reducing the time and configuration burden required to run complex analyses in teams where bioinformatics talent is expensive and unevenly distributed [Drylab].
| Market lens | Public evidence | Interpretation |
|---|---|---|
| Regulated biopharma research software | Drylab targets "highly regulated biopharmaceutical and biomedical organizations" [The Business Times, October 2026] | Indicates enterprise R&D budgets rather than consumer or academic-only tooling |
| Bioinformatics workflow automation | Drylab says it automates dataset discovery, custom model training, and one-click bioinformatics pipelines [Drylab] | Suggests overlap with workflow orchestration and analysis software spend |
| Secure on-premise or local AI research tooling | Drylab says it offers a desktop mode where data does not leave the user’s machine, and enterprise security features including MFA and AES-256 [Drylab] | Points to demand from institutions with data-governance constraints |
| Applied multi-omics and translational research | Drylab cites biomarker discovery and patient stratification as use cases [Drylab] | Connects the product to discovery and clinical-adjacent research workflows |
The useful read-through from these segments is that Drylab is likely selling into a budget pool assembled from several existing categories, not creating a wholly new one. That can help early adoption if the ROI is concrete, but it can also lengthen sales cycles because ownership may sit across R&D, bioinformatics, IT, and compliance functions.
Adjacent and substitute markets are material here. A buyer could address parts of the same problem with internal bioinformatics teams, conventional workflow software, CRO partnerships, point solutions for omics analysis, or broad enterprise AI platforms adapted for life sciences, rather than adopting a purpose-built autonomous research agent system [Drylab] [The Business Times, October 2026]. That means Drylab’s market is not simply "AI for science" in the abstract. It is the narrower slice where automation, security posture, and deployability inside institutional environments justify replacing existing combinations of people, scripts, and software.
Regulation and macro conditions cut both ways. The positive side is that security and compliance burdens can favor vendors that are built for controlled environments from the start, and Drylab’s public positioning leans heavily in that direction, including local-data handling and enterprise security claims [Drylab]. The constraint is that regulated buyers tend to validate cautiously, and claims around autonomous agents in scientific research may face higher scrutiny than generic productivity tools, especially where outputs could influence expensive wet-lab or clinical-adjacent decisions [The Business Times, October 2026] [drylab-bio.github.io].
A final nuance is that the company’s public benchmark work, DRYLAB-Bench, sits near an emerging biosecurity discussion around model-guided mutation design, which broadens the policy context even if it does not yet prove commercial demand [drylab-bio.github.io]. For market formation, that can be a double signal: more attention on scientific AI capabilities, but also more institutional caution around governance, model evaluation, and acceptable use. For a young company, that usually means the addressable market may be real before it is fully liquid.
No independent source found -- This section relies mainly on company materials, one named funding article, and an under-review project page; no independent third-party market sizing report was available in the provided sources.
The Competitive Field
MIXED Drylab AI is positioning itself less against a single named software rival than against the existing workflow stack inside translational research, where scientists move among bioinformatics tools, local scripts, and internal compute environments rather than buy one autonomous research layer [Drylab] [The Business Times, October 2026].
The public record is thin on named direct competitors, which matters here because it limits how precisely the field can be mapped from independent evidence. What is visible is Drylab's chosen segment: regulated biopharma and biomedical research teams that want AI-assisted analysis without pushing sensitive data into a fully external workflow [The Business Times, October 2026] [Drylab]. In practice, that places the company across three competitive lanes at once: incumbent bioinformatics workflows composed of many separate tools, adjacent enterprise AI platforms that could add life-science modules, and internal teams at pharmaceutical companies or research institutes that may prefer to build agentic research tooling on their own infrastructure [Drylab] [The Business Times, October 2026].
The current edge, to the extent public evidence supports one, is product packaging around regulated deployment rather than a demonstrated distribution moat. The Business Times reports that Drylab can be deployed inside a pharmaceutical company's or research institution's infrastructure, while the company says its desktop mode keeps data on the user's machine and that its security layer includes MFA and AES-256 encryption [The Business Times, October 2026] [Drylab]. That combination is commercially relevant because data handling is often the first procurement obstacle in biomedical research, but the durability is still uncertain: security posture can be copied by better-capitalized platforms, and there is no public evidence yet of proprietary datasets, exclusive channel partners, or named lighthouse customers that would make the wedge harder to replicate [Drylab] [Crunchbase].
The main exposure is that Drylab appears to be selling into organizations that already have reasons to resist category consolidation. The company says it integrates 800-plus bioinformatic tools and automates multi-omics analysis, model training, and pipeline execution [Drylab]. If that claim holds in practice, the platform could reduce workflow fragmentation; if it does not, incumbent point tools and internal computational biology teams retain an advantage because they already sit inside validated scientific processes and do not require a buyer to trust a newer agentic layer with end-to-end execution [Drylab] [The Business Times, October 2026]. A second exposure is buyer proof: The Business Times reports paying customers across pharmaceutical companies, cancer centers, biomedical research institutes, and universities, but no names are disclosed, so the market cannot yet see whether Drylab is winning in large enterprise accounts, smaller innovation budgets, or research pilots that may not expand cleanly [The Business Times, October 2026].
Over the next 18 months, the most plausible competitive outcome is not category winner-take-all but segmentation by deployment preference. Drylab is a likely winner if regulated research groups continue to prefer in-environment deployment and local data control over general-purpose cloud AI workflows, because that is the one point the public record supports with some specificity [The Business Times, October 2026] [Drylab]. Drylab is also the most plausible loser if large platform vendors or internal biopharma engineering teams absorb the same orchestration features into existing environments faster than startups can establish account-level trust, especially since no public source yet shows Drylab owning a distribution channel, a benchmark standard with field adoption, or a named account base that competitors would struggle to displace [Crunchbase] [drylab-bio.github.io, Under review at ICLR 2027].
Opportunity
PUBLIC
If Drylab AI executes, the prize is not a better bioinformatics tool, it is a place inside the core research workflow of regulated life science organizations, where software budgets, switching costs, and downstream decision value can all compound quickly once trust is established [The Business Times, October 2026] [Drylab].
The headline opportunity is to become the operating layer through which biopharma and biomedical teams run AI-assisted research, especially in settings where security, deployment control, and workflow automation matter more than raw model novelty. That outcome is reachable, at least in outline, because the public record already shows three ingredients that usually have to arrive separately: a product framed around end-to-end workflow automation, deployment options that can live inside a customer's own infrastructure, and early evidence of paying customers across multiple institution types and regions by October 2026 [Drylab] [The Business Times, October 2026]. In practical terms, if a platform can help researchers discover datasets, train custom models, run bioinformatics pipelines, and keep sensitive data within enterprise guardrails, it moves closer to infrastructure than to point software [Drylab] [The Business Times, October 2026].
The upside case still depends on execution, but the shape of the wedge is visible. Drylab says the platform supports both a web app and a desktop app, and that its Desktop App mode keeps data on the user's machine, while The Business Times reports that the product can also be deployed within a pharmaceutical company's or research institution's infrastructure [Drylab] [The Business Times, October 2026]. For buyers in regulated research, that architecture can matter as much as model quality, because procurement often turns on where data sits, who controls it, and how much manual workflow overhead remains.
| Scenario | What happens | Catalyst | Why it's plausible |
|---|---|---|---|
| Secure research-system-of-record | Drylab becomes the default AI workbench for sensitive biomedical analysis inside pharma, cancer centers, and research institutes | A visible enterprise deployment pattern around on-premise or customer-controlled infrastructure, paired with security features such as MFA and AES-256 in Drylab Vault | The company already positions security as a product feature, and outside coverage says the platform can be deployed inside customer infrastructure for regulated organizations [Drylab] [The Business Times, October 2026] |
| Multi-omics workflow standard | Drylab wins by compressing weeks of analysis work into a much faster, repeatable workflow for both no-code researchers and expert bioinformaticians | A category-tipping use case such as biomarker discovery or patient stratification proving repeatable ROI inside one customer segment | The product claims are aimed at exactly those workflows, including multi-omics analysis, automated code generation, and one-click pipelines [Drylab] |
| Biosecurity and evaluation moat | Drylab uses DRYLAB-Bench and related evaluation work to become a trusted benchmark and tooling layer for agentic biology systems | External adoption of its benchmark by labs, model developers, or enterprise buyers assessing risk and performance | The company has publicly associated itself with a benchmark focused on agentic LLM mutation design and real wet-lab measurements, which could support trust and technical authority if adoption follows [drylab-bio.github.io] |
What compounding looks like here is less a consumer-style network effect and more an enterprise research flywheel. If one team uses the platform to automate dataset discovery, model training, and pipeline execution, the next sale can start with a clearer ROI story and a lower implementation burden, especially if deployment works inside the customer's existing environment [Drylab] [The Business Times, October 2026]. In regulated settings, trust features can also compound. A product that meets internal security expectations can move from one lab or program into adjacent workflows faster than a tool that requires exceptions each time.
There is also a product-surface compounding effect if Drylab's claims hold in practice. The company says it integrates 800+ bioinformatic tools, supports no-code and expert users, and offers use cases from biomarker discovery to patient stratification and prediction [Drylab]. That creates a plausible path where each successful workflow broadens the set of users inside an account, from bench-adjacent scientists to computational biology teams, which in turn can raise retention and wallet share. The reported presence of paying customers across the US, Europe, and Asia by October 2026 suggests at least an early signal that the product thesis is not limited to a single geography or institution type [The Business Times, October 2026].
The size of the win is harder to pin down because the available source set does not include a named market-sizing report or a directly comparable public peer in the provided materials. Even so, a concrete upside frame is possible. If Drylab became a trusted infrastructure layer for AI-driven biomedical research inside large life science organizations, its value would likely be judged less like a narrow workflow utility and more like strategic research software with deep enterprise embed. With a disclosed US$3 million seed round in October 2026, the company is still at a stage where category ownership remains open [The Business Times, October 2026]. A plausible bull case is a future strategic-software outcome measured in the high hundreds of millions of dollars, or more, if one of the scenarios above matures into broad enterprise standardization (scenario, not a forecast). That is an inference from the category position implied by the product and customer claims, not a market-backed valuation model.
No independent source found -- This section relies materially on company claims from Drylab's website and one reported funding article from The Business Times; the upside framing is analytical and partially inferred rather than independently corroborated.
Sources
Publicly reported
[Crunchbase] Drylab AI | Crunchbase Company Profile & Funding | https://www.crunchbase.com/organization/drylab-ai
[The Business Times, October 2026] AI agent platform Drylab AI raises US$3 million seed round from Vertex Ventures and PatSnap | https://www.businesstimes.com.sg/startups-tech/startups/ai-agent-platform-drylab-ai-raises-us3-million-seed-round-vertex-ventures-and-patsnap
[LinkedIn] Drylab | https://www.linkedin.com/company/drylab-ai
[Drylab] AI Bioinformatics Platform for Biomedical Research | Drylab | https://thedrylab.com/
[LinkedIn, retrieved 2026] Nhi Nguyen - Bosch Global Software Technologies Vietnam | LinkedIn | https://www.linkedin.com/in/nhi-nguyen-2a2520371/
[drylab-bio.github.io, Under review at ICLR 2027] DRYLAB-Bench - Benchmarking the Biosecurity Risk of Agentic LLM Mutation Design | https://drylab-bio.github.io/
Articles about Drylab AI
- Drylab AI's Autonomous Agents Are Running Bioinformatics Pipelines for Pharma — The seed-stage startup is betting that autonomous AI, not just copilot tools, can accelerate biomarker discovery and patient stratification in highly regulated environments.