Location
Ottawa and Toronto, Canada - Hybrid
Other Canadian locations - Remote
About the team
Our AI team develops machine-learning solutions across forecasting, segmentation, explainability, and optimization for industries including Consumer Packaged Goods, Retail, and Life Sciences. We combine rigorous science with real business problems, and our work is deployed to enterprise customers worldwide.
Demand forecasting is central to the Kinaxis platform. We are building a dedicated data-science and AI research engine to deliver forecasting that is accurate, self-calibrating, explainable, and cost-effective at enterprise scale.
About the role
We are seeking a Senior Staff Applied Scientist with deep expertise in time-series and demand forecasting to set the scientific direction for our forecasting platform. This is a hands-on senior individual-contributor role with no direct reports. You will define the science, establish technical standards, and guide multiple engineering and data-science teams to deliver the roadmap.
You will own the scientific strategy and measurement framework used to improve forecasting accuracy, robustness, and credibility. You will also influence long-term technical direction as the platform scales to millions of series and daily forecasting.
Vacancy Status
This is an existing job vacancy
What you will do
- Set the forecasting science agenda. Define the research roadmap and translate complex business goals into clear ML problems, measurable success criteria, and technical priorities.
- Lead through influence. Provide technical direction, mentor senior scientists and ML engineers, review designs, and strengthen scientific judgment across teams.
- Advance forecasting capabilities. Guide evaluation and development across classical time-series, machine- learning, probabilistic, and pretrained forecasting approaches, including hierarchical and intermittent demand, covariates, cold starts, segmentation, explainability, and domain adaptation.
- Establish scientific rigor. Define accuracy metrics, benchmarks, quality gates, backtesting standards, and reproducible experiment-to-validation practices.
- Shape architecture for scale. Partner with architects, MLOps, and platform and data-infrastructure teams to ensure solutions can support millions of series and daily forecasting.
- Communicate and align. Explain scientific trade-offs to technical and executive audiences and build alignment across accuracy, scalability, cost, operational simplicity, and commercial priorities.
- What Success Looks Like
- Within six months, a fixed and reproducible accuracy benchmark is in place, teams report against it, and the research roadmap is aligned with engineering and product leadership.
- Within the first year, benchmark-verified accuracy improvements have shipped to production, and teams independently design and evaluate experiments using the standards you established.
What we are looking for
- Deep, hands-on expertise in time-series or demand forecasting, including classical statistical methods, modern ML approaches, and deep-learning or foundation-model methods.
- Strong knowledge of backtesting, hierarchical and intermittent demand, covariates, forecast evaluation, probabilistic and quantile forecasting, calibration, and uncertainty.
- Master's degree in computer science, statistics, operations research, or a related quantitative field; Ph.D. preferred.
- 8+ years of experience developing, testing, debugging, and optimizing complex ML solutions in Python, with strong software-engineering practices.
- Experience moving research and proofs of concept into maintainable production systems with measurable impact.
- Proficiency with distributed data tooling such as Spark, cloud platforms such as GCP or Azure, containers, and distributed computing.
- Practical experience with model explainability, reproducibility, evaluation standards, and ML quality gates.
- Demonstrated ability to shape technical direction across multiple teams, influence roadmap and investment decisions, mentor others, and remain hands-on.
- Strong written and verbal communication, including the ability to align technical teams and executive stakeholders around a scientific strategy.
- Nice to Have
- Publications, patents, open-source contributions, industry benchmarks, or production systems with demonstrated impact.
- Experience with pretrained time-series models such as Chronos, TabPFN-TS, or TimesFM, and with LLM-assisted configuration or analytics tooling.
- Supply chain or manufacturing knowledge, including demand planning, S&OP, replenishment, or optimization, and experience forecasting hundreds of thousands to millions of series.
- Experience productizing data-science capabilities for non-experts and building automated configuration, calibration, or accuracy assistants.
#Senior; #LI-EM1
Work With Impact: Our platform directly helps companies power the world’s supply chains. We see the results of what we do out in the world every day, when we see store shelves stocked, when medications are available for our loved ones, and so much more.
Work with Fortune 500 Brands: Companies across industries trust us to help them take control of their integrated business planning and digital supply chain. Some of our customers include Lockheed Martin, Unilever, P&G, ExxonMobil, Cisco and more.
Social Responsibility at Kinaxis: Our Diversity, Equity, and Inclusion Committee weighs in on hiring practices, talent assessment training materials, and mandatory training on unconscious bias and inclusion fundamentals. Sustainability is key to what we do and we’re committed to a long-term net-zero operations strategy. We are involved in our communities and support causes where we can make the most impact.
People matter at Kinaxis and here are some of the perks and benefits we offer, which may vary by location and employee:
- Flexible vacation and Kinaxis Days (company-wide days off)
- Flexible work options
- Physical and mental well-being programs
- Regularly scheduled virtual fitness classes
- Mentorship programs, training, and career development
- Recognition programs and referral rewards
- Hackathons
For more information, visit the Kinaxis website at www.kinaxis.com or the company’s blog at http://blog.kinaxis.com.
