ServiceNow Research
ServiceNow Research is the artificial intelligence research division of ServiceNow, the publicly listed American enterprise-software company headquartered in Santa Clara, California. The research division is anchored in Montreal and Toronto with the senior research team that joined ServiceNow through the January 2021 acquisition of Element AI, the Yoshua-Bengio-affiliated Canadian AI research startup. ServiceNow Research operates an active publication program at major AI conferences, contributed to the open-research BigCode Project co-leading the StarCoder and StarCoder2 open-weights code-foundation-model line, and develops the Now Assist generative AI capabilities embedded across the ServiceNow product line. As of April 2026, ServiceNow Research is one of the principal enterprise-AI research divisions, with the dual academic-research and commercial-product orientation distinctive among Fortune-500 enterprise-software organizations.
At a glance
- Founded: 2021 as ServiceNow Research, integrating the Element AI research team acquired January 2021. Element AI was originally founded in 2016 by Yoshua Bengio, Jean-Francois Gagne, and other Canadian AI researchers.
- Status: Research division of ServiceNow, a publicly listed company on the New York Stock Exchange (NYSE: NOW).
- Funding: Operates within ServiceNow's research-and-development budget. ServiceNow's annual research-and-development investment exceeds $2 billion as of fiscal year 2025 to 2026.
- CEO / Lead: Bill McDermott, ServiceNow Chief Executive Officer (since 2019). Research division leadership operates under ServiceNow's Chief Technology Officer organization.
- Other notable leadership: Senior research leadership includes Sarath Chandar, Joao Monteiro, Issam Laradji, Pierre-Luc Bacon, and other senior research scientists; the broader research organization is anchored in Montreal and Toronto with research-collaboration relationships across Canadian AI institutions including Mila and the Vector Institute.
- Open weights: Yes, partial. ServiceNow Research has co-led the BigCode Project and contributed to the open-weights StarCoder and StarCoder2 model line releases.
- Flagship outputs: BigCode Project co-leadership; StarCoder and StarCoder2 open-weights code-foundation-model line; Now Assist generative AI capabilities; published research at NeurIPS, ICML, ICLR, ACL.
Origins
ServiceNow Research traces its lineage to Element AI, the Montreal-based AI research startup founded in 2016 by Yoshua Bengio (the University of Montreal professor and Mila Scientific Director who shared the 2018 Turing Award with Geoffrey Hinton and Yann LeCun), Jean-Francois Gagne, Nicolas Chapados, and other Canadian AI researchers. Element AI built a substantial research-and-applied-AI organization focused on enterprise applications of deep learning across approximately 500 staff in Montreal, Toronto, and other global offices.
ServiceNow's January 2021 acquisition of Element AI consolidated the research team into ServiceNow Research, with the Montreal and Toronto offices retained as the principal research-organization locations. Bengio departed Element AI and continued his academic role at Mila and the University of Montreal, while many of the senior Element AI research scientists transitioned into ServiceNow Research roles. The acquisition was characterized in industry coverage as ServiceNow's principal AI-research-talent positioning among enterprise-software incumbents.
The 2022 to 2024 period saw substantial commercial AI product development under the Now Assist brand, with generative-AI capabilities integrated into the ServiceNow Workflow Automation, IT Service Management, Customer Service Management, and HR Service Delivery product lines. The ServiceNow Now Platform AI capabilities are positioned as the principal differentiation for enterprise-software customers, with the research-division contribution providing the foundation-model and applied-research depth that ServiceNow's enterprise sales-organization can position against Salesforce Einstein, Microsoft Copilot, SAP Joule, and other competitor enterprise-AI offerings.
The 2023 to 2026 period has seen continued open-research output, including the foundational role in the BigCode Project, the open-research collaboration that produced StarCoder (May 2023), StarCoder2 (February 2024), and continued open-weights code-foundation-model releases. ServiceNow Research's collaboration with Hugging Face, NVIDIA, and other BigCode partners has been characterized as one of the principal open-research-cooperation programs in the foundation-model ecosystem.
Mission and strategy
ServiceNow Research's mission is to advance the foundation-model and applied-AI research that enables ServiceNow's enterprise-AI product line, with continued contribution to the open-research community as a research-credibility anchor. The strategy combines three threads. First, foundation-model research with focus on code-generation, agentic systems, and structured-output reasoning, contributing to the BigCode Project open-research collaboration. Second, applied-AI research that translates into Now Assist generative-AI capabilities across the ServiceNow product line. Third, research-collaboration relationships with Canadian academic AI institutions (Mila, Vector Institute, University of Montreal, University of Toronto) that anchor research-talent recruitment and academic credibility.
The competitive premise is that enterprise-software companies that build genuine AI research capability will produce structurally better AI products than competitors that integrate third-party foundation models without in-house research depth.
Models and products
- BigCode Project co-leadership. Open-research collaboration co-led with Hugging Face, NVIDIA, and other research peers. Released StarCoder (May 2023) and StarCoder2 (February 2024) open-weights code-foundation-model line.
- StarCoder and StarCoder2. Open-weights code-generation foundation models. Trained on permissively licensed code from The Stack training dataset (also a BigCode-Project release).
- Now Assist. Generative AI capabilities embedded across the ServiceNow product line including IT Service Management, Customer Service Management, HR Service Delivery, and Workflow Automation.
- Published academic research. Active publication program at NeurIPS, ICML, ICLR, ACL, and other major AI conferences. Research areas include foundation-model training, code generation, agentic AI systems, and reinforcement learning.
- Research-cooperation relationships. With Mila, Vector Institute, University of Montreal, University of Toronto, and other Canadian academic AI institutions.
Distribution channels are predominantly through ServiceNow's enterprise-software product line for the commercial Now Assist offering, and through Hugging Face and GitHub for the open-research and open-weights releases.
Benchmarks and standing
ServiceNow Research's evaluation framework combines academic publication metrics (paper count and citation impact at NeurIPS, ICML, ICLR), open-weights model adoption (StarCoder and StarCoder2 download counts on Hugging Face), and commercial product metrics (Now Assist adoption rates, ServiceNow AI revenue contribution).
The StarCoder and StarCoder2 model line has been characterized in industry coverage as one of the principal open-weights code-foundation-model releases globally, with research-community traction comparable to peer code-foundation-model releases from Meta AI (Code Llama), DeepSeek (DeepSeek-Coder), and Alibaba Qwen (Qwen2.5-Coder).
Leadership
As of April 2026, ServiceNow Research's leadership includes:
- Bill McDermott, ServiceNow Chief Executive Officer (since November 2019). Former CEO of SAP. Public face for ServiceNow's overall AI strategy.
- Senior research leadership anchored in Montreal and Toronto with research scientists including Sarath Chandar, Joao Monteiro, Issam Laradji, Pierre-Luc Bacon, and other senior researchers from the Element AI lineage.
- Element AI legacy talent that transitioned into ServiceNow Research roles after the January 2021 acquisition.
Funding and backers
ServiceNow Research operates within ServiceNow Corporation's research-and-development budget. ServiceNow's annual research-and-development investment exceeds $2 billion as of fiscal year 2025 to 2026. ServiceNow's market capitalization in 2025 to 2026 has been in the multi-tens-of-billions of dollars range on the New York Stock Exchange. The company's strong cash flow from enterprise-software subscription revenue supports continued AI research investment.
Industry position
ServiceNow Research occupies a distinctive position among enterprise-software AI research divisions, with the Element AI research-team lineage, the BigCode Project open-research-collaboration role, and the Now Assist commercial-product integration. Industry coverage has characterized ServiceNow as the enterprise-software incumbent with the most credible AI research organization, with the Yoshua Bengio research lineage and the StarCoder open-weights releases as the principal validating data points.
Competitive landscape
- Salesforce AI Research, Microsoft AI, Adobe Research, SAP, Oracle. Direct enterprise-software AI research peers.
- BigCode Project, Hugging Face, NVIDIA Research. Open-research-cooperation peers and partners.
- Mila, Vector Institute. Canadian academic AI research peers and research-cooperation partners.
- Meta AI / FAIR, DeepSeek, Alibaba Qwen / DAMO. Code-foundation-model competitors with open-weights releases.
Outlook
- The continued cadence of BigCode Project open-research output through 2026 to 2027.
- Continued Now Assist generative AI capability expansion across the ServiceNow product line.
- The continued research-collaboration relationships with Canadian academic AI institutions.
- Continued enterprise-AI commercial growth tied to ServiceNow's overall product-line traction.
Sources
- ServiceNow Research. Research division reference.
- BigCode Project. Open-research collaboration.
- StarCoder on Hugging Face. Open-weights code-foundation-model line.
- Element AI Wikipedia. Predecessor organization.
- ServiceNow NYSE listing NOW. Public-market financial reporting.