SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
ServiceNow AI Research is seeking a Principal Research Scientist to lead research initiatives on enterprise agentic AI. You will work on research questions that most AI labs do not have access to, with a rare vantage point: observing how agents actually behave inside complex enterprise environments at scale.
You will lead research initiatives focused on how agents accumulate and reuse tacit knowledge, learn and improve continuously, and on their resilience, safety, and data privacy. These are the real obstacles to enterprise agentic deployment today, surfacing in operational reality well before appearing in academic literature.
While much of the AI and NLP community concentrates on core agent capabilities, this team explores the dimensions that determine real adoption: safety mechanisms, threat detection, privacy guarantees, and responsible AI practices. You will have the freedom to publish and advance the state of the art, with the rare advantage of testing hypotheses against real deployments rather than benchmarks.
You will influence research direction across the organization, collaborate with product and engineering teams, and help shape how enterprise AI systems are built and deployed responsibly. The role offers a bilingual work environment in Montreal with flexible work arrangements.
**Requirements:**
- PhD in computer science, machine learning, or related field, or equivalent research experience
- Eight or more years of post-PhD research experience, with significant portion in industrial setting or direct contact with deployed systems
- Established publication record at top-tier venues (NeurIPS, ICML, ICLR, ACL, EMNLP, or equivalents) with contributions recognized as influential in your field
- Demonstrated track record of setting research agenda, not only executing: you have chosen problems before the community turned to them and can explain why
- Experience providing scientific leadership to cross-disciplinary teams or projects bringing together researchers and engineers
- Strong software development skills: write, maintain, and optimize high-quality research code; understand what it takes for results to survive production deployment
- Advanced use of agentic tools in your own research practice and willingness to advance team practices
- Demonstrated ability to influence technical decisions beyond your own team, particularly with product and engineering teams
- Excellent communication skills, including ability to explain complex technical work to non-specialist audiences and executives
**Nice to have:**
- Expertise in agent memory and tacit knowledge, continual learning, safety and alignment, security and threat detection, privacy, or agent evaluation
- Experience with AI systems operating in real enterprise environments with compliance, governance, and scale constraints
- Contributions to open source software, datasets, or evaluation benchmarks used by the community
- Scientific service roles: program committees, editorial work, workshop organization
- Academic collaborations or supervision of doctoral students
- French fluency or interest in learning it for bilingual Montreal environment
