Merck & Co. (MSD outside the US and Canada)

Ortent Life Sciences Radar, research beta. Evidence to 6 October 2026, rule set v0.4.3. Scope: AI and computation in drug discovery (T1), and research-lab digitalisation and scientific data systems (T2). Each ID in brackets links to its online source, and the source index at the end of this brief lists them.

Disclosure: Andrew Wyatt, Ortent's principal, was Chief Growth Officer at Sapio Sciences until April 2026. Software vendors, including Sapio and its competitors, appear in this brief only where public sources name them.

Status: research beta. Ortent has reviewed this brief; it has not been independently validated. It summarises activity identified in public sources. Search coverage varies by company and is set out under Coverage limits, and zero identified items does not mean no activity. It does not predict buying intent or measure operational maturity. For this radar, process and CMC development are grouped with manufacturing and excluded from the research-lab theme (T2); such work appears as context where it helps.

Summary

The reviewed sources identify six AI-discovery collaboration commitments involving MSD: two first disclosed in the year to 4 October 2025 and four in the year to 5 October 2026 [>MSD-E08, >E10, >E14, >E16, >E20, >E22]. MSD company stories and a research web page describe in-house foundation models, LLM-based AI agents and R&D data platforms, with no stated start dates [>MSD-E03, >E04, >E12, >E13]. In company lab digitalisation (T2), the searched sources contain one commitment, first disclosed in the year to 4 October 2025: a Benchling collaboration to bring MSD's regulated bioanalytical workflows onto one platform [>MSD-E24]. We identified no T2 commitment change in the year to 5 October 2026. The 6 October recruitment snapshot contains 5 relevant R&D requisition postings across the two themes (3 T1, 2 T2); vacancy status is not recorded for any of them [>MSD-R001, >R002, >R004, >R005, >R006]. A sixth, a lab automation post in pharmaceutical development labs, is process development and outside the themes under the v0.4.3 boundary [>MSD-R003]. These observations identify programmes and functions for further investigation; they do not establish procurement opportunity, deployment completion or net hiring.

Announced commitments

MSD issued one of the six T1 items jointly with its partner [>MSD-E16]. The other five were announced first by the partner, with an MSD executive quoted in four [>MSD-E10, >E14, >E20, >E22]. Only a Turbine executive is quoted in the Turbine release [>MSD-E08]. MSD posted a summary of each of these five on its merck.com business development and licensing (BD&L) pages on the day of the partner release [>MSD-E08, >E10, >E14, >E20, >E22].

The Benchling release quotes Roy Helmy, Associate Vice President, Regulated Bioanalytics at Merck Research Laboratories. The release states intent ("to implement"), not completed deployment. We found no BD&L item or other dated MSD confirmation. A 6 October 2026 internship posting in the Regulated Bioanalytical Sciences team names Benchling as that team's laboratory information management system [>MSD-R022]. The posting has no date, so it is not treated as a dated confirmation of the collaboration. The release was found in a search top-up run after the reviewers' searches, through a T2 vendor search [>MSD-E24]. Regulated bioanalysis is an R&D lab under clarification C42-11, so the collaboration is in T2.

The Illumina Atlas "is already in build", in the release's words. MSD's T1 tag rests on the stated training of its own AI/ML models; the data access element is context under rule AR7a, not T2 [>MSD-E14]. Mayo Clinic data access is likewise context for T2, and the T1 tag rests on the stated target identification scope [>MSD-E16]. Protillion's own lab-in-the-loop platform is partner capability under AR7a, so it is not counted in T2 [>MSD-E22].

A collaboration with Boston Consulting Group to use AI and generative AI for the discovery and assessment of novel drug targets was announced on 30 May 2024, before both windows, and is not counted in either year [>MSD-E01].

By event date, we counted two T1 initiatives with a qualifying commitment change in W0 (5 October 2024 to 4 October 2025), Turbine and Variational AI, and four in W1 (5 October 2025 to 5 October 2026), the Illumina Billion Cell Atlas, Mayo Clinic, Infinimmune and Protillion. By first disclosure, the T1 count is also two in W0 and four in W1. T2 is one in W0 (Benchling) and zero in W1, in both views [>MSD-E24]. No milestone updates were disclosed in either theme. Existing capabilities first disclosed numbered one in T1 and one in T2 in W0 [>MSD-E03, >E04], and three in T1 and none in T2 in W1 [>MSD-E12, >E13, >E15]. These have no stated start date and are not counted as changes. No requisition took theme both, so clarification C42-19 does not affect the recruitment split below.

Two of the v0.4.2 clarifications marked borderline would change these counts if read the other way. Under the opposite reading of C42-4, the Merck Digital Sciences Studio (MDSS) Montreal opening and its second cohort would count as T1, giving T1 counts of 3 in W0 and 5 in W1 in both views [>MSD-E06, >E18]. Under the opposite reading of C42-2, the Evaxion vaccine in-licence would count as T1, giving 3 in W0 and 4 in W1 in both views [MSD-E11]. With both read the other way, T1 would be 4 in W0 and 5 in W1. C42-12 does not apply to MSD. T2 is one in W0 and zero in W1 under every reading.

Coverage is partial in every count cell. The company press archive, annual reports and results materials were searched completely in both years. Partner newswires, consortium and venture announcements were searched only in part, as were academic announcements for T2. A search top-up ran most of the web searches the shared budget had cut, but some still did not run (see Coverage limits). The coverage table marks those families as not comparable between the years, so no comparison between years is made.

The MDSS accelerator and the Tempus collaboration are recorded as context, not commitments, under clarifications C42-4 and C42-8 (see Context).

Identified recruitment

On 6 October 2026 one reviewer screened 105 postings on the jobs.merck.com and jobs.msd.com career sites and the other screened 38 on the Workday careers site; 33 were screened by both. We classified 110 requisitions from their duties. We observed 5 relevant requisition postings in the R&D theme class (3 T1, 2 T2) [>MSD-R001, >R002, >R004, >R005, >R006]. Neither careers site shows an openings field, so vacancy status is "not recorded" for every posting under clarification C42-23, and none is counted as ambiguous. A further 4 are postdoctoral research roles (2 T1), 34 are internships or co-ops (4 T1, 6 T2), 27 are enterprise roles, 13 are plant, quality or manufacturing roles and 27 are out of scope. These are requisition postings listed on one day, not hires, headcount or a trend.

The T1 postings describe AI and computational roles in discovery: - an Associate Director, AI/ML Engineering, in Cambridge, Massachusetts, to build AI/ML tools and pipelines, "particularly transformer-based Foundation Models", in the Data, AI and Genome Sciences (DAGS) function, which the posting describes as taking "an AI/ML-first approach to improving target and biomarker discovery, validation and selection" [>MSD-R002] - a Senior Specialist in the Cheminformatics group of Modeling and Informatics, Discovery Chemistry, in Cambridge, Massachusetts, to apply AI/ML and cheminformatics workflows in the early target discovery space and improve hit identification. The posting describes Modeling and Informatics as "approximately 50 computational chemists, cheminformaticians, and data scientists" [>MSD-R006] - a Senior Manager, AI, in the MSD R&D China Innovation Collaboration Center in Shanghai and Beijing, to build "advanced AI systems for drug evaluation and development", with experience in structure prediction, target-ligand binding prediction and ADMET prediction [>MSD-R004]

The T2 postings describe scientific data roles: - a Lead Semantic Data Engineer in Prague and Hyderabad, to design knowledge graph architectures that integrate scientific data such as omics, process and analytical data, and to set a knowledge graph roadmap "aligned with enterprise data product strategy" [>MSD-R001] - a Senior Technical Product Manager in Prague, in the Research Data and Analytics Platform of Global Research IT, to set the roadmap of "the integration platform connecting scientific systems globally" [>MSD-R005]

A Lab Automation / Controls Engineer in Rahway, New Jersey, in the DSCS Digital Technologies team of Development Sciences and Clinical Supply, is to work towards "automated and connected laboratories through the utilization and deployment of laboratory SCADA systems and orchestration platforms" in pharmaceutical development labs [>MSD-R003]. These are process development labs, so under the v0.4.3 boundary the posting is out of scope; it was a T2 posting under v0.4.2.

Two DAGS and vaccine discovery postdoctoral roles are tagged T1 but sit outside the R&D theme count: computational pipelines for target and biomarker discovery from genetics, proteomics and health record data [>MSD-R007], and AI/ML and computational protein design for vaccine antigens [>MSD-R009]. Ten internships or co-ops are tagged to the themes. The four T1 roles cover AI/ML computational toxicology, data science, and modelling and informatics [>MSD-R019, >R027, >R040, >R041]. The six T2 roles cover discovery sample management automation, an AI-enabled knowledge repository for lab automation scripts, semantic data engineering, lab automation engineering, and two regulated bioanalytical lab projects [>MSD-R017, >R021, >R012, >R035, >R022, >R023]. The two bioanalytical projects are in T2 under clarification C42-11 [>MSD-R022, >R023].

Enterprise postings include an AI Platform Engineer in Prague, in Advanced Scientific Compute within Global Research IT, to run on-premise GPU and HPC infrastructure for AI model training and inference. Its duties name no discovery activity, so it is an enterprise role under clarification C42-21 [>MSD-R048]. A DAGS senior data scientist role building foundation models for predictive immune biomarkers is out of scope as biomarker work [>MSD-R093].

Posting dates come from page fields that look like refresh dates, so they are not reliable.

No LinkedIn company insights were collected for MSD.

Technology mentions in recruitment

Mentions are classified from the exact posting wording. They do not establish installation, an incumbent contract or company-wide use. The table records 103 MSD mentions: 8 explicit use, 6 integration or environment, 1 planned implementation, 10 required experience, 4 preferred experience, 70 alternatives and 4 unclear. - Named as existing systems: Benchling, which a regulated bioanalytical internship posting names as the team's laboratory information management system, used to capture "assay workflows and experimental data" [>MSD-R022]. Jira, used to keep the integration platform's release process on track [>MSD-R005]. Outside the themes, the AI Platform Engineer posting states that training and inference workloads run "primarily utilizing Slurm, Nvidia Triton and Docker containers" on on-premise GPU hardware [>MSD-R048], and a Hyderabad data engineering posting builds pipelines "on Databricks and AWS" [>MSD-R058]. - Planned implementation: the process development lab automation engineer, a posting outside the themes from v0.4.3, is to help deploy "laboratory SCADA systems and orchestration platforms". No product is named in that sentence [>MSD-R003]. - Integration responsibilities: knowledge graph platforms [>MSD-R001], an integration platform connecting scientific systems [>MSD-R005], high-throughput automated expression and screening platforms [>MSD-R009], data transfer tools to integrate instruments with LIMS [>MSD-R017], and GitHub repositories and lab automation systems [>MSD-R021]. No lab system product is named. - Required experience: Python, Torch, AWS, MLFlow, Docker, GitHub and GitHub Actions [>MSD-R002]; AWS [>MSD-R001]. - Preferred experience: Claude Code, Databricks and GCP [>MSD-R002]. - Lists of examples: Ignition, DeltaV and WinCC OA as SCADA or orchestration software, in the process development posting [>MSD-R003]; MuleSoft, Kong, Apache Kafka and AWS [>MSD-R005]; Amazon Neptune, RDF and SPARQL [>MSD-R001]; OEChem, RDKit and DeepChem, and Pipeline Pilot, KNIME and Spotfire [>MSD-R006]; PyTorch, TensorFlow and JAX [>MSD-R004]; ESM, AlphaFold and ProteinMPNN [>MSD-R009]; "Benchling, LIMS or other electronic laboratory notebook platforms" [>MSD-R022]. - Unclear: three internship postings refer to "electronic notebook recording" without naming a system [>MSD-R028, >R031, >R042].

No category (ELN, LIMS, cloud) is recorded as explicit use in the theme postings. Benchling is the only lab informatics product named as an existing system, and only in one internship posting for one regulated bioanalytical team [>MSD-R022]. Separately, Benchling's own release of June 2025 announced a collaboration with MSD's regulated bioanalytics function [>MSD-E24].

Deployment evidence

• A company story of 3 February 2025 says that "a team of high-performing AI assistants" based on large language models helps researchers at Merck Research Labs "focus on critical drug discovery and development". It gives no start date [>MSD-E03].
• A company story of 9 January 2026 says: "We recently developed two foundation models which uncover patterns in disease to find better drug targets, allow faster molecular design and rapidly test small molecules". The models are not named, and no date is stated [>MSD-E13].
• KERMT, a small-molecule model "Developed in collaboration with Nvidia" and trained on more than 11 million molecules, was first described in a preprint by MSD authors on 14 October 2025. A company story of 19 March 2026 calls it open source [>MSD-E12]. We found no distinct release announcement, so it is a disclosed capability, not a counted release.
• A Healthcare Brew article of 11 December 2025 quotes an MSD senior director saying KERMT and TEDDY "have versions deployed directly to scientists' desktops". This is media-reported. It is recorded in the KERMT row note and is not treated as a deployment claim [>MSD-E12].
• The KERMT company story quotes a senior director of data science, from that interview, as saying AI "is sometimes cutting timelines by 30% or more". No basis or programme is given, so we do not treat it as a measured result [>MSD-E12].
• The merck.com research page on data science and AI says MSD is building scalable data platforms "at every level in R&D". This is a category statement with no system named. The page carries a publication date of 28 January 2025 in its metadata and a modification date of 21 September 2026, so the sentence may date from a later edit [>MSD-E04].
• An Illumina release of 13 January 2026 names Iya Khalil as vice president and head of Data, A.I. & Genome Sciences. No appointment date is stated [>MSD-E15].
• We found no adoption or benefit metric for any of these items in the reviewed sources.

Context

Leadership: the DAGS head is named in a partner release with no appointment date [>MSD-E15]. A media report of 23 July 2026 says Gaurav Gupta has been appointed Executive Vice President and Chief Information & Digital Officer, to assume the role on 1 September 2026, succeeding Dave Williams. His stated remit covers the company's information technology and digital agenda, including AI and enterprise data and technology platforms. No discovery remit is stated, so this is context under clarification C42-7 [>MSD-E23].

Enterprise programmes: on 22 April 2026 MSD and Google Cloud announced a multi-year partnership, valued at up to $1 billion, that "will deploy an agentic platform" across R&D, manufacturing, commercial and corporate functions. No discovery or lab-system scope is stated, so it is context under clarification C42-9 [>MSD-E21]. A June 2025 release says an internal generative AI platform for clinical study reports cut the time to a first draft from 180 hours to 80; this is clinical writing AI and is context [>MSD-E05].

Under clarification C42-8, the Tempus collaboration is context, not T1. On 3 March 2026 Tempus and MSD reported an expanded, multi-year collaboration under which MSD will use Tempus' de-identified data and Lens platform. The MSD quote places the AI work in precision biomarkers, resistance mechanisms and combinations, not target identification or candidate selection [>MSD-E17]. The earlier Tempus agreement has no date in the reviewed sources.

Under clarification C42-4, the Merck Digital Sciences Studio accelerator is context (venture). Merck Canada opened MDSS in Montreal on 20 June 2025 for seven startups, each funded through the Merck Global Health Innovation Fund and CQDM or Merck Canada [>MSD-E06]. A second cohort of 12 startups was announced on 13 March 2026, each to receive $100,000 in total from the Merck Global Health Innovation Fund, The Venture Collective and CQDM [>MSD-E18]. Both releases say MDSS targets "preventive health, platforms for biologics research and development (R&D) and 3D translational disease models". AI appears as one example of digital technology, not as a stated discovery scope.

Under clarification C42-2, the in-licence of Evaxion's vaccine candidate EVX-B3, with up to $592 million in milestones, is context. Evaxion describes itself as developing AI-Immunology powered vaccines, but no AI work, platform or model access for MSD is stated [MSD-E11]. Two target discovery deals, with Mestag and Quotient, state no AI or computation in their scope and are excluded [>MSD-E02, >E19].

Workforce and sites: in July 2025 MSD approved a restructuring programme expected to "eliminate certain administrative, sales and R&D positions", with cumulative pretax costs of about $3.0 billion and annual savings of about $1.7 billion, substantially realised by the end of 2027 [>MSD-E07]. A media report of 11 September 2025 quotes an MSD statement that it plans "to discontinue discovery research operations in the UK" and no longer plans to occupy the Belgrove House site at King's Cross; the article reports that 125 scientific jobs will be cut [>MSD-E09]. Neither source links these to AI in discovery or lab digitalisation, and we identified no qualifying pause or cancellation in the two themes in the searched sources, which now include a T2-specific negative search. That is a search result, not evidence that none occurred.

Coverage limits

• The shared web-search budget ran out during this batch. A later search top-up of ten web searches ran the T2 negative-event search; the T2 capacity, leadership, partnership, acquisition, venture and consortium searches; one regulatory programme search for both themes; one IHI consortium search; and one academic AI-discovery collaboration search. It found the Benchling collaboration [>MSD-E24] and no other qualifying item. These searches still did not run: the T2 open release search, a general T1 capacity search, a T2 academic search, the UK Biobank, SGC and AISB membership checks, the NVIDIA newsroom search and the Global Health Innovation Fund portfolio check. Ortent keeps a record of the exact queries.
• Three top-up results were not read because their sites refused robots.txt: a Revvity Signals case study on "Merck" informatics, a MarketScreener item on "Merck" management appointments and a World Pharmaceuticals item on two "Merck" AI drug discovery collaborations. Each could concern Merck KGaA, a separate company. They are not used.
• Results slide decks for Q3 2024 to Q1 2025 were not linked on the results event pages, although the results cells are marked complete.
• Partner newswires, consortium and corporate venture announcements were searched only in part in both years, and academic collaborations only for T1. There was no systematic wire-site search, the NVIDIA newsroom was not searched, and the Merck Global Health Innovation Fund portfolio page was not fetched.
• Business Wire, GlobeNewswire, sec.gov, Fierce Pharma, Pharmaceutical Technology, both Tempus sites, Barchart, the IHI site, parallelparliament.co.uk, msdcareers.com and one BCG page blocked access or could not be reached. Endpoints disallows AI crawlers in robots.txt and was not fetched. The Infinimmune and Protillion releases were read through BioSpace copies and the Tempus release through a 1stOncology copy; the 10-K filings were read from company-hosted copies.
• The London Discovery Centre statement rests on a media article. No company copy was found, and the statement date is unknown [>MSD-E09].
• The research page date comes from page metadata, not a visible date [>MSD-E04].
• The Turbine release page states that Turbine is backed by the MSD Global Health Innovation Fund, with no date. No investment row was made, because no date can be established [>MSD-E08].
• A report that the DAGS head later moved to another company was seen only in a search result from Endpoints. It is not verified and is not used.
• TEDDY, named in Healthcare Brew as an MSD genomics foundation model, has no bundled release announcement and no row.
• Patents are not among the searched signal types. Publication counts are raw Europe PMC query counts and are not events.
• Recruitment: the two reviewers used different careers front ends, and neither sitemap is guaranteed to list every open requisition. Only the 110 requisitions classified are in the table.
• Two independent reviewers searched MSD, and the final rows are their source-verified union plus one row from the search top-up. All six counted T1 commitments were found by both reviewers, but overlap does not show that both did not miss the same events. The T2 Benchling commitment was found only by the top-up, which shows that the reviewers' searches had missed at least one qualifying item [>MSD-E24].

Two questions to investigate (Ortent's interpretation)

1. Lab automation and scientific data roles sit in several units. A DSCS Digital Technologies engineer in Rahway is to help deploy lab SCADA and orchestration platforms in pharmaceutical development labs; this is process development, outside T2 under the v0.4.3 boundary, but relevant to the interfaces [>MSD-R003]. A product manager in Global Research IT is to lead the integration platform connecting scientific systems [>MSD-R005], and a semantic data engineer is to build knowledge graphs aligned with enterprise data product strategy [>MSD-R001]. A regulated bioanalytical team names Benchling as its own lab system [>MSD-R022], and Benchling announced a collaboration with MSD's regulated bioanalytics function in June 2025 [>MSD-E24]. MSD's research page says it is building data platforms "at every level in R&D" [>MSD-E04], the Google Cloud agentic platform is to be deployed across R&D [>MSD-E21], and a new Chief Information & Digital Officer was due to take up the role on 1 September 2026 [>MSD-E23]. What is the remit of each unit for lab systems and scientific data? Who decides on new lab automation and data systems, and how are the interfaces between these units and the enterprise programme governed?
2. MSD has signed external AI-discovery agreements, each quoting a discovery chemistry or biologics leader from Merck Research Laboratories [>MSD-E10, >E20, >E22]. It also describes foundation models it has developed, and DAGS is recruiting to build foundation model pipelines [>MSD-E12, >E13, >MSD-R002]. A Modeling and Informatics role includes evaluating "new technology and software from commercial sources, open innovation, or external collaborations" [>MSD-R006]. MSD posted each partner-announced deal on its BD&L pages [>MSD-E08, >E10, >E14, >E20, >E22]. Who holds decision rights over licensing an external AI discovery platform rather than building one, and does such licensing run through BD&L, procurement or the research functions? How is funding for these decisions set within the restructuring programme and its savings target [>MSD-E07]?

Sources for the main claims

• MSD-E08 Turbine and MSD collaboration: https://turbine.ai/news/turbine-msd-2025/
• MSD-E10 Variational AI collaboration with MSD (press kit PDF): https://connect.variational.ai/hubfs/Press%20Kit/PR_Variational%20AI%20Enters%20Collaboration%20with%20Merck.pdf
• MSD-E14 Illumina Billion Cell Atlas: https://www.illumina.com/company/news-center/press-releases/2026/fda84c92-b4b3-4691-a402-35555abe8605.html
• MSD-E16 MSD and Mayo Clinic agreement: https://www.merck.com/news/merck-and-mayo-clinic-announce-new-research-and-development-collaboration-to-support-ai-enabled-drug-discovery-and-precision-medicine/
• MSD-E20 Infinimmune collaboration with MSD (BioSpace copy): https://www.biospace.com/press-releases/infinimmune-enters-collaboration-with-merck-to-discover-novel-human-antibody-therapeutics
• MSD-E22 Protillion collaboration with MSD (BioSpace copy): https://www.biospace.com/press-releases/protillion-biosciences-announces-drug-discovery-collaboration-with-merck
• MSD-E24 Benchling collaboration with MSD on vaccine bioanalysis: https://www.benchling.com/news/benchling-collaboration-with-merck-vaccine-bioanalysis
• MSD-E01 BCG collaboration (msd.com BD&L item): https://www.msd.com/bdl_item/bcg-announces-genai-collaboration-with-msd-tradename-of-merck-co-inc-rahway-n-j-usa/
• MSD-E06 MDSS opens in Canada: https://www.merck.ca/en/newsroom/merck-digital-sciences-studio-opens-in-canada-to-accelerate-biopharmaceutical-innovation/
• MSD-E18 MDSS second cohort: https://www.merck.com/news/merck-digital-sciences-studio-expands-with-second-innovation-class-to-advance-biopharma-and-digital-health-progress-in-north-america/
• MSD-E12 KERMT company story: https://www.merck.com/stories/our-ai-model-kermt-is-helping-to-advance-drug-discovery/
• MSD-E13 Five ways story: https://www.merck.com/stories/5-ways-were-transforming-artificial-intelligence-into-impact/
• Other event rows and all requisitions: URLs and archived copies in the MSD tables. No SRC source files are cited for MSD.

Source index

Each ID in brackets links to the online source for that row. Job posting links come from a snapshot of the company job site on 6 October 2026; postings that have closed since are marked and no longer linked. Where a company site blocked access, the link goes to the copy that was read.