Expanded preview entry for Microsoft Discovery brings new enterprise-grade, agentic AI capabilities for analysis and improvement groups.
Over the previous yr, we’ve made important progress with Microsoft Discovery by working carefully with analysis and improvement (R&D) organizations. Immediately, we’re sharing how these efforts are translating into actual momentum for purchasers and companions, whereas additionally increasing preview entry to Microsoft Discovery. This subsequent section displays what we’ve discovered as we proceed to broaden entry to enterprise-grade, agentic AI capabilities for R&D. The Microsoft Discovery platform continues to evolve with new capabilities, expanded companion interoperability, and a rising set of outcomes with real-world scientific outcomes and engineering transformation. We consider what comes subsequent can meaningfully change how R&D groups function and empower them to attain extra.
The period of agentic AI for analysis and improvement
Agentic AI opens a brand new chapter for R&D the place autonomous agent groups, guided by human experience, carry out the core analysis and engineering duties in a redefined agentic loop. Specialised brokers can purpose on high of huge quantities of organizational and public-domain data, create hypotheses on an expanded search house, check and validate these hypotheses at scale, analyze the outcomes, and feed conclusions into iterative loops. Empowering science and engineering consultants with agentic AI has the potential to reshape the way forward for science and engineering, enabling organizations to guide boldly within the new Frontier R&D period.
This elementary shift requires a deep transformation that encompasses each technological and organizational challenges. Scientific discovery has at all times been outlined by ambition and the relentless pursuit of what comes subsequent—a extra sustainable materials, a cleaner supply of power, a more practical therapy. However for a lot of R&D groups the toughest work can start after an thought reveals promise. Turning ideas into outcomes requires repeated improvement cycles that contain reformulating candidates as new datasets emerge, re-engineering current supplies to satisfy evolving regulatory and efficiency necessities, or adjusting designs when efficiency, yield, or manufacturability fall quick. As R&D grows extra advanced, tooling should evolve to assist shut the gap between what researchers and engineers wish to pursue and what they’ll virtually ship.
Earlier generations of AI provided incremental aid by way of sooner search and higher retrieval, however lacked the deeper reasoning that genuinely advanced, multi-disciplinary science calls for. Tradeoffs throughout price, efficiency, yield, compliance, and timelines should be revisited repeatedly as improvement progresses. However the convergence of large-scale reasoning fashions, agentic AI architectures, and high-performance cloud infrastructure has created a real alternative to rethink how R&D work will get achieved—not solely to enhance current processes on the margins, however to assist groups iterate sooner and transfer from speculation to candidate improvement to end result with larger confidence.
When Microsoft Discovery was launched in personal preview final yr, it was an early expression of that risk: an agentic AI platform purpose-built for R&D, bringing collectively the reasoning depth and collaborative intelligence that advanced, real-world R&D requires. The response from engineers and researchers throughout life sciences, chemistry and supplies science, physics, semiconductors, and different fields made clear that the necessity was actual and the method was proper.
The Microsoft Discovery platform
Microsoft Discovery is an extensible platform that brings collectively agentic orchestration, superior reasoning, a graph-based data basis, and high-performance computing. It helps drive the three ideas outlined in Determine 1 for efficient agentic discovery—enabling agent empowerment, discovery loop automation, and high quality at scale. As a result of it’s constructed on Microsoft Azure’s enterprise cloud infrastructure, Microsoft Discovery is designed to function throughout the safety, compliance, transparency, and governance frameworks used to handle delicate real-world R&D environments.
Brokers are geared up with a broad vary of digital, bodily, and analytical instruments used throughout R&D. This contains in silico experimentation environments similar to high-performance compute (HPC) clusters, specialised giant quantitative fashions (LQMs) and brokers, and potential future integration with quantum capabilities as they grow to be relevant to industrial R&D. It additionally permits interoperability with bodily labs, facilitating the lab process technology and even direct operation with robotics, lab instrumentation, and Web of Issues (IoT)-enabled units that brokers can function beneath human oversight.
At the center of Microsoft Discovery is the Discovery Engine that mimics the scientific methodology the place specialised brokers purpose over giant quantities of information, generate hypotheses, and validate them in a posh tree throughout an unlimited search house. The Discovery Engine connects proprietary analysis knowledge with exterior scientific literature—not solely to retrieve remoted info however to purpose throughout conflicting theories, experimental outcomes, and domain-specific assumptions in a method that displays how science truly works. This contextual depth is what separates Microsoft Discovery from general-purpose AI instruments and permits the platform to perform as a real considering companion throughout the total arc of a analysis program.
Constructed-in governance controls assist be certain that agent pushed analysis stays aligned with strategic priorities, safety and compliance requirements, and security necessities. These programs present centralized administration, audit trails, and checkpoints that assist preserve reliability as agentic throughput grows. The platform is extensible by design which permits integration with current enterprise instruments and property, companion options, and open-source fashions. Integration with Microsoft 365, Microsoft Foundry, and Microsoft Material permits organizations to interoperate throughout enterprise brokers, enterprise knowledge, and institutional data.
Actual-world impression of Microsoft Discovery
Beforehand we shared how a group of Microsoft researchers leveraged superior AI fashions and HPC instruments from Microsoft Discovery to determine a novel, non-PFAS, immersion datacenter coolant prototype in about 200 hours. We’re excited to share a couple of examples of how prospects have been utilizing the platform throughout preview.
Syensqo
A world chief in superior supplies and specialty chemical substances, Syensqo is advancing a daring, multi-year transformation of its know-how panorama to speed up data-driven science, superior simulation, and AI-enabled discovery. Constructing on early success with Microsoft Discovery, Syensqo is now scaling these capabilities enterprise-wide to unlock larger scientific and enterprise impression. This subsequent section focuses on modernizing R&D data foundations, increasing entry to scalable, cost-efficient, cloud-based compute, and establishing a unified working mannequin that brings collectively knowledge, high-performance computing, and rising agentic AI to energy the way forward for innovation.
As Microsoft Discovery workflows gained momentum, Syensqo expanded its ambition to scale these capabilities throughout each R&D and industrial organizations, unlocking new alternatives for end-to-end innovation. This evolution is enabling groups to unify scientific and enterprise datasets, scale simulation environments consistent with more and more advanced improvement wants, and combine engineering workflows inside a related digital ecosystem. Collectively, these developments are establishing a powerful, future-ready basis to speed up innovation-led progress—from early-stage discovery by way of engineering and large-scale formulation.
To comprehend this imaginative and prescient, Syensqo is advancing its science and industrial knowledge and simulation platforms on Azure. By centralizing important datasets inside a ruled, enterprise-grade knowledge spine and increasing Microsoft Discovery workflows onto extremely scalable cloud compute, the corporate is establishing a contemporary, standardized working mannequin for innovation. This shift permits extra seamless collaboration, helps superior analytics and simulation at scale, and lays the groundwork for next-generation, AI-powered workflows throughout precedence analysis and innovation (R&I) domains.
We’re coming into a brand new section of our partnership with Microsoft, targeted on scaling AI brokers throughout analysis, gross sales and advertising to drive near-term progress. By connecting buyer demand to scientific improvement and again to market execution, agentic AI is enabling sooner cycles, sharper prioritization, and tangible impression on income progress and enterprise efficiency.”
—Mike Radossich, Chief Government Officer (CEO), Syensqo
GigaTIME
Trendy oncology more and more will depend on understanding tumors not solely by look, however by the organic indicators that form cell habits, immune response, and therapy outcomes. GigaTIME addresses this want through the use of AI to deduce spatially resolved tumor microenvironment indicators from routine hematoxylin and eosin (H&E) pathology slides. This method makes insights similar to immune infiltration, checkpoint context, and tumor proliferation extra accessible at scale with out the associated fee and throughput constraints of experimental assays. GigaTIME and its outputs inside Microsoft Discovery are meant for analysis use solely. They aren’t a medical gadget and should not meant for scientific prognosis, therapy, prevention, or patient-management selections.
The impression of GigaTIME will increase when its outputs are embedded into actual analysis workflows. Inside Microsoft Discovery, digital multiplex immunofluorescence (mIF) predictions transfer past standalone visualizations and grow to be inputs to ongoing scientific reasoning. Spatial phenotypes will be generated persistently throughout cohorts, localized to single cell context, and related to supporting proof similar to literature, biomarkers, and downstream endpoints. This permits researchers to interpret outcomes systematically, query assumptions, and refine organic hypotheses over time.
Microsoft Discovery helps this work in a method that’s reproducible, scalable, and ruled finish to finish. GigaTIME can be utilized alongside extra fashions, knowledge sources, and instruments inside a shared setting that helps iteration, comparability, and validation. Fairly than accelerating a single analytical step, Discovery helps a full discovery loop—the place spatial biology informs hypotheses, hypotheses information validation, and outcomes feed the subsequent cycle of studying with readability and confidence.
Be taught extra in regards to the GigaTIME and Microsoft Discovery integration to see how digital mIF outputs are utilized inside Microsoft Discovery for oncology R&D.
PhysicsX
PhysicsX, a pacesetter in physics AI for industrial engineering and manufacturing, is partnering with Microsoft to convey agentic engineering into manufacturing by way of Microsoft Discovery. On the core of this collaboration is the PhysicsX platform—combining Giant Physics Fashions and AI-native workflows to ship near-real-time simulation by inference throughout the total engineering lifecycle.
Built-in into Discovery’s agentic setting, the PhysicsX platform permits engineers to maneuver past sequential, solver-driven workflows and discover considerably bigger design areas, evaluating hundreds of manufacturable candidates in days, with out compromising bodily constancy.
The collaboration is already delivering impression at Microsoft Floor. Confronted with tightly coupled constraints throughout thermal efficiency, acoustics, and kind issue, the Floor engineering group used the PhysicsX platform by way of Discovery to reimagine their cooling fan design course of. What beforehand required weeks of simulation and handbook setup is now compressed into days. Discovery brokers orchestrate the technology, analysis, and optimization of hundreds of geometries, surfacing high-performing, production-ready designs for validation.
The result’s a step change in engineering productiveness: sooner iteration, broader design-space protection, and extra assured decision-making. The method is now being prolonged throughout extra elements within the Floor portfolio.
Engineering continues to be constrained by workflows constructed for the pre-AI period. This partnership modifications that. PhysicsX’s frontier physics AI fashions, mixed with Microsoft Discovery’s agentic orchestration and Azure infrastructure, give engineers the flexibility to discover design areas that had been beforehand out of attain—on the pace and scale that trendy industrial improvement calls for.
—Jacomo Corbo, CEO, PhysicsX
Synopsys
Synopsys is a pacesetter in digital design automation (EDA), laptop aided engineering (CAE) instruments, and mental property (IP), and performs a central function within the design and improvement of essentially the most advanced chips and programs for the main semiconductor and programs corporations of the world.
Synopsys and Microsoft have been partnering since 2019, serving to pioneer software-as-a-service (SaaS) fashions on Microsoft Azure. Synopsys additionally launched the primary Silicon Copilot in collaboration with Microsoft and is constant that journey by leveraging Microsoft Discovery to roll out options for chip design.
The semiconductor trade is going through an unprecedented set of challenges—demand for prime efficiency chips is rising exponentially, complexity of sustainable, power-efficient chip design, and a important scarcity of expert engineering. Agentic programs may also help mitigate these challenges whereas accelerating design cycles.
Synopsys agentic AI stack with multi-agent workflows constructed on AgentEngineer™ know-how, supported by Microsoft Discovery, have outlined a brand new paradigm for the trade.
Chip design sits on the intersection of utmost complexity and outsized impression—precisely the place AI could make the largest distinction. By bringing collectively Synopsys’ AI‑pushed design management with Microsoft Discovery, we’re enabling agentic AI to redefine semiconductor engineering workflows, unlock step‑perform productiveness good points, and speed up the subsequent period of know-how innovation.
—Ravi Subramanian, Chief Product Administration Officer, Product Administration & Markets Group, Synopsys
A rising ecosystem
Microsoft Discovery works with an increasing ecosystem of companions providing built-in instruments and specialised experience.
Increasing what is feasible for R&D
Increasing the preview marks an vital step in making agentic AI accessible to a broader set of R&D organizations. Microsoft Discovery displays our perception that the subsequent technology of scientific progress can come from programs that mix human experience with AI that may purpose, plan, and act at scale.
We stay up for partnering with organizations that wish to rethink how discovery occurs and to assist form the way forward for enterprise R&D.
For organizations trying to get began with Microsoft Discovery remember to evaluate the technical documentation to know necessities, onboarding stipulations, and infrastructure issues.
Microsoft Discovery is obtainable in preview. Options, availability, integrations, and efficiency traits described on this publish might change previous to, or with out, basic availability and should not commitments. Statements about future capabilities (together with any potential quantum integration) are forward-looking and topic to vary. Buyer and inside outcomes described mirror particular workflows and knowledge; particular person outcomes will differ.
