{"id":793670,"date":"2024-07-09T16:12:12","date_gmt":"2024-07-09T23:12:12","guid":{"rendered":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/?post_type=msr-group&p=793670"},"modified":"2026-09-18T16:59:44","modified_gmt":"2026-09-18T23:59:44","slug":"m365-research","status":"publish","type":"msr-group","link":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/group\/m365-research\/","title":{"rendered":"M365 Research"},"content":{"rendered":"
\n\t
\n\t\t
\n\t\t\t\"M365\t\t<\/div>\n\t\t\n\t\t
\n\t\t\t\n\t\t\t
\n\t\t\t\t\n\t\t\t\t
\n\t\t\t\t\t\n\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t\t\t\n\t\t\t\t\t\t\t\n\n

M365 Research<\/h1>\n\n\n\n

Delivering Agentic AI systems that combine high-quality outcomes with enterprise-scale efficiency<\/p>\n\n\t\t\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t<\/div>\n\t\t<\/div>\n\t<\/div>\n<\/section>\n\n\n\n\n\n

Advancing enterprise-grade Agentic AI<\/h2>\n\n\n\n
\n
\n

M365 Research advances enterprise-grade Agentic AI so Microsoft Copilot can become a trusted and skilled digital coworker embedded in the flow of enterprise work. We develop the foundational capabilities and systems that help agents understand people, goals, processes, and organizational context; execute complex work reliably; learn safely from validated experience; and operate with the performance and economics required at global scale.<\/p>\n\n\n\n

Enterprise work spans people, documents, meetings, messages, applications, processes, permissions, and teams. Agents operating in this environment must assemble the right context, retain useful experience, reason across multiple steps, use tools safely, recognize failures, and remain under appropriate human control.<\/p>\n\n\n\n

These capabilities also increase latency, infrastructure demand, and cost. M365 Research therefore treats agent quality, reliability, performance, and efficiency as one connected systems problem, spanning enterprise context, workflow traces, governed learning, realistic evaluation environments, validated repair, and efficient serving infrastructure.<\/p>\n<\/div>\n\n\n\n

\n
\n

Can Agentic AI understand and execute enterprise work well enough to become indispensable\u2014and can it do so with the performance, efficiency, and cost required at global scale?<\/p>\n<\/blockquote>\n<\/div>\n<\/div>\n\n\n\n

Two complementary research pillars<\/h2>\n\n\n\n

M365 Research connects two complementary lines of work: improving what enterprise agents can accomplish reliably and improving the systems that make those capabilities responsive and economically sustainable at scale. Together, they allow us to study the agent and the systems that support it as a connected research problem.<\/p>\n\n\n\n

\"Two<\/figure>
\n

Enabling enterprise agentic experiences<\/h3>\n\n\n\n

Increasing agent capability can increase inference cost, context size, and latency. M365 Research works across that tension, advancing the quality and reliability of agentic systems while improving the efficiency of the models, infrastructure, and execution stack that support them.<\/p>\n\n\n\n

Together, these pillars enable enterprise agentic experiences: digital coworkers that can understand complex enterprise context, take reliable action, and operate responsively and economically at global scale.<\/p>\n<\/div><\/div>\n\n\n\n

\n
\n

Agentic Innovation<\/h4>\n\n\n\n

Quality. Reliability. Trust.<\/p>\n\n\n\n

Advances reasoning, planning, memory, skills, multi-agent collaboration, evaluation, and governed self-correction so agents can consistently deliver high-quality enterprise work outcomes.<\/p>\n\n\n\n\n\n

High-quality agentic work depends on understanding the environment in which work occurs: people, teams, documents, meetings, messages, projects, business processes, permissions, and organizational norms. We develop governed and persistent context that enables agents to remember relevant information, learn recurring workflows and procedures, acquire and apply reusable skills, maintain freshness and provenance, and operate within appropriate boundaries.<\/p>\n\n\n\n

Memora<\/a> | LegoMem<\/a> | ACON<\/a><\/p>\n\n\n\n\n\n

Agent progress depends on evidence that reflects actual enterprise outcomes. We develop task-grounded evaluation methods, realistic and privacy-preserving synthetic environments, and data-quality analysis that distinguishes failures caused by models, retrieval, tools, environments, data, or graders. This work helps identify low-signal training examples before compute is spent and supports shared quality bars for determining whether improvements will generalize to real product workloads.<\/p>\n\n\n\n\n\n

Enterprise agents must reason across complex workflows, recognize uncertainty, detect failures, identify their causes, and recover without silently producing poor outcomes. We develop diagnosis-to-repair loops that localize failures in execution traces, propose bounded changes to prompts, tools, or orchestration, and validate those changes against regression checks before deployment.<\/p>\n\n\n\n

AgentRx<\/a> | AutoSaddler<\/a> | Algorithmist<\/a><\/p>\n\n\n\n\n\n

We bring foundational capabilities into real productivity and engineering workflows where research can improve reliability, productivity, and efficiency. These settings include enterprise knowledge work, software modernization, workload knowledge, cloud operations, incident diagnosis, and account recovery.<\/p>\n\n\n\n

FLASH<\/a><\/p>\n\n\n<\/div>\n\n\n\n

\n

Efficient AI<\/h4>\n\n\n\n

Performance. Cost. Scale.<\/p>\n\n\n\n

Advances inference optimization, latency and cost efficiency, and hardware-software co-design so Agentic AI can operate responsively and economically across the Microsoft 365 ecosystem.<\/p>\n\n\n\n\n\n

GenAI workloads differ substantially in context size, latency needs, quality requirements, safety requirements, caching potential, and execution patterns. We develop workload-aware serving techniques spanning traffic partitioning, endpoint tailoring, scheduling, batching, routing, load balancing, and configuration selection.<\/p>\n\n\n\n

GAUSS<\/a><\/p>\n\n\n\n\n\n

AI performance depends on the interaction among models, serving configurations, compute kernels, memory behavior, communication collectives, hardware topology, and workload shape. We co-design these layers around real Microsoft 365 workloads to improve throughput and useful work per GPU-second.<\/p>\n\n\n\n

LeanAttention<\/a> | Capillary Kernels<\/a><\/p>\n\n\n\n\n\n

Reasoning models and long-running agents accumulate context, reuse memory, call tools, branch across steps, and operate over extended periods. We develop KV-cache lifecycle management, context-footprint reduction, asynchronous and throughput-oriented serving, adaptive compute, and workflow-level cost-and-quality measurement so complex agentic experiences remain responsive and economically sustainable.<\/p>\n\n\n\n

Prompt Compression<\/a> | AutoSaddler<\/a> | LegoMem<\/a><\/p>\n\n\n\n\n\n

We use agents to help explore large configuration, scheduling, and kernel-design spaces while keeping researchers accountable for problem framing, validation, and interpretation. We also build reusable research knowledge and evaluation assets that help distinguish promising technical results from improvements that will not translate into customer or business outcomes.<\/p>\n\n\n<\/div>\n<\/div>\n\n\n\n

<\/div>\n\n\n","protected":false},"excerpt":{"rendered":"

Delivering Agentic AI systems that combine high-quality outcomes with enterprise-scale efficiency M365 Research advances enterprise-grade Agentic AI so Microsoft Copilot can become a trusted and skilled digital coworker embedded in the flow of enterprise work. We develop the foundational capabilities and systems that help agents understand people, goals, processes, and organizational context; execute complex work […]<\/p>\n","protected":false},"featured_media":1185645,"template":"","meta":{"msr-url-field":"","msr-podcast-episode":"","msrModifiedDate":"","msrModifiedDateEnabled":false,"ep_exclude_from_search":false,"_classifai_error":"","msr_group_start":"","footnotes":""},"research-area":[13556,13558,13547],"msr-group-type":[243694],"msr-locale":[268875],"msr-impact-theme":[],"class_list":["post-793670","msr-group","type-msr-group","status-publish","has-post-thumbnail","hentry","msr-research-area-artificial-intelligence","msr-research-area-security-privacy-cryptography","msr-research-area-systems-and-networking","msr-group-type-group","msr-locale-en_us"],"msr_group_start":"","msr_detailed_description":"","msr_further_details":"","msr_research_lab":[],"related-researchers":[{"type":"user_nicename","display_name":"Saravan Rajmohan","user_id":41039,"people_section":"Leadership","alias":"saravar"},{"type":"user_nicename","display_name":"Gbola Afonja","user_id":42846,"people_section":"Team","alias":"gafonja"},{"type":"user_nicename","display_name":"Chetan Bansal","user_id":31394,"people_section":"Team","alias":"chetanb"},{"type":"user_nicename","display_name":"Agrim Bari","user_id":44276,"people_section":"Team","alias":"agrimbari"},{"type":"user_nicename","display_name":"Srikant Bharadwaj","user_id":41644,"people_section":"Team","alias":"srbharadwaj"},{"type":"user_nicename","display_name":"Krishna Chintalapudi","user_id":32577,"people_section":"Team","alias":"krchinta"},{"type":"user_nicename","display_name":"Camille Couturier","user_id":40111,"people_section":"Team","alias":"cacoutur"},{"type":"user_nicename","display_name":"Mayukh Das","user_id":41140,"people_section":"Team","alias":"mayukhdas"},{"type":"user_nicename","display_name":"Dhruv Deshmukh","user_id":44203,"people_section":"Team","alias":"ddeshmukh"},{"type":"user_nicename","display_name":"Prathamesh Deshpande","user_id":43323,"people_section":"Team","alias":"deshpandepra"},{"type":"user_nicename","display_name":"Hang Dong","user_id":39687,"people_section":"Team","alias":"hangdong"},{"type":"user_nicename","display_name":"Zhiyu Fan","user_id":44213,"people_section":"Team","alias":"fanzhiyu"},{"type":"user_nicename","display_name":"Jay Gala","user_id":43965,"people_section":"Team","alias":"t-jaygala"},{"type":"user_nicename","display_name":"Mirian Hipolito Garcia","user_id":40483,"people_section":"Team","alias":"mirianh"},{"type":"user_nicename","display_name":"Huseyin Atahan Inan","user_id":40426,"people_section":"Team","alias":"huinan"},{"type":"user_nicename","display_name":"Sam Ade Jacobs","user_id":43503,"people_section":"Team","alias":"samjacobs"},{"type":"user_nicename","display_name":"Liqun Li","user_id":43104,"people_section":"Team","alias":"liqli"},{"type":"user_nicename","display_name":"Minghua Ma","user_id":41218,"people_section":"Team","alias":"minghuama"},{"type":"user_nicename","display_name":"Daniel Eduardo Madrigal Diaz","user_id":40480,"people_section":"Team","alias":"danielmad"},{"type":"user_nicename","display_name":"Ankur Mallick","user_id":42441,"people_section":"Team","alias":"ankurmallick"},{"type":"user_nicename","display_name":"Anjaly Parayil","user_id":41215,"people_section":"Team","alias":"aparayil"},{"type":"user_nicename","display_name":"Bo Qiao","user_id":37848,"people_section":"Team","alias":"boqiao"},{"type":"user_nicename","display_name":"Xiaoting Qin","user_id":43008,"people_section":"Team","alias":"xiaotingqin"},{"type":"user_nicename","display_name":"Si Qin","user_id":42006,"people_section":"Team","alias":"siqin"},{"type":"user_nicename","display_name":"Saravan Rajmohan","user_id":41039,"people_section":"Team","alias":"saravar"},{"type":"user_nicename","display_name":"Victor Ruehle","user_id":41027,"people_section":"Team","alias":"virueh"},{"type":"user_nicename","display_name":"Robert Sim","user_id":36650,"people_section":"Team","alias":"rsim"},{"type":"user_nicename","display_name":"Gagan Somashekar","user_id":43416,"people_section":"Team","alias":"gsomashekar"},{"type":"user_nicename","display_name":"Renee St. Amant","user_id":43080,"people_section":"Team","alias":"reneestamant"},{"type":"user_nicename","display_name":"Rujia Wang","user_id":42549,"people_section":"Team","alias":"rujiawang"},{"type":"user_nicename","display_name":"Molly Xia","user_id":41943,"people_section":"Team","alias":"mollyxia"},{"type":"user_nicename","display_name":"Fangkai Yang","user_id":41425,"people_section":"Team","alias":"fangkaiyang"},{"type":"user_nicename","display_name":"Jue Zhang","user_id":41212,"people_section":"Team","alias":"juezhang"},{"type":"user_nicename","display_name":"Xuchao Zhang","user_id":42045,"people_section":"Team","alias":"xuchaozhang"},{"type":"user_nicename","display_name":"Pu Zhao","user_id":38886,"people_section":"Team","alias":"puzhao"},{"type":"user_nicename","display_name":"Qingwei Lin \u6797\u5e86\u7ef4","user_id":33318,"people_section":"Team","alias":"qlin"}],"related-publications":[507653,510482,547407,594046,597184,597430,632667,635433,710380,710392,715375,723331,747061,747079,764191,771604,780454,781984,785137,790529,807226,823309,830107,837859,838957,841759,846043,846301,846307,846901,864054,864876,865395,869661,873306,879069,881646,882387,909372,910413,915804,931692,939681,939924,939993,940899,945099,945279,945306,950790,957018,975756,976098,978456,986508,990273,994119,1011312,1016619,1016919,1017021,1018440,1020141,1020150,1023294,1026390,1026447,1031928,1031946,1041231,1041966,1068603,1078539,1078572,1078584,1078599,1078614,1078641,1078659,1083105,1083138,1083144,1090611,1091175,1091574,1100199,1100211,1100217,1100226,1100241,1103526,1122570,1127547,1128276,1129749,1129848,1131444,1134645,1134647,1134649,1134651,1135010,1135014,1136484,1139560,1139563,1149140,1150229,1152852,1152856,1152860,1152863,1152925,1153960,1153962,1153976,1154194,1155073,1155076,1155079,1155432,1155884,1156474,1156484,1161825,1163586,1169496,1169499,1173094,1173174,1173235,1175802,1176044,1178912,1181563,1185705],"related-downloads":[],"related-videos":[],"related-projects":[1186747,1186810,1155944,1110357,1017939,866259,855579,507611],"related-events":[],"related-opportunities":[],"related-posts":[793088,842272,939198,964494,987321,997686,1013043,1029624,1085448,1095585,1102680,1104735,1116780,1163539,1168790],"tab-content":[],"msr_impact_theme":[],"_links":{"self":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-group\/793670","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-group"}],"about":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/types\/msr-group"}],"version-history":[{"count":190,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-group\/793670\/revisions"}],"predecessor-version":[{"id":1186815,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-group\/793670\/revisions\/1186815"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/media\/1185645"}],"wp:attachment":[{"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/media?parent=793670"}],"wp:term":[{"taxonomy":"msr-research-area","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/research-area?post=793670"},{"taxonomy":"msr-group-type","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-group-type?post=793670"},{"taxonomy":"msr-locale","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-locale?post=793670"},{"taxonomy":"msr-impact-theme","embeddable":true,"href":"https:\/\/cm-edgetun.pages.dev\/en-us\/research\/wp-json\/wp\/v2\/msr-impact-theme?post=793670"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}