What You Can Expect

  • Invaluable takeaways from world-class experts assessing the emerging and longer-term future of AI’s impact on planning, forecasting, and employment
  • Superb opportunities to interact with our speakers and network with your peers
  • Ample session time for Q&A
  • Complimentary copies of the new Foresight guidebook on AI
  • A serene setting at one of North America’s finest hotels and conference centers— 2017 and 2018 TripAdvisor Certificate of Excellence recipient

Program agenda, subject to change

Who Should Attend?

  • Forecasters and planners
  • Managers in all areas of operations, marketing, and finance
  • Academics, researchers and data scientists addressing AI
  • Anyone who wants to gain insight into AI’s effects on forecasting and planning


Duncan Klett
Co Founder & Fellow

Using AI/ML to deliver a smarter supply chain

A good demand forecast is better than no forecast at all, and AI/ML methods can certainly provide a better forecast, but it’s what you do with that forecast that counts.  In this session, learn how to apply AI/ML models for:

  • Strategy and segmentation,
  • Managing change, and
  • Mitigating risks.

Each of these areas enable organizations to respond to supply chain variability more effectively to achieve business success.

Stephan Kolassa
Data Science Expert, SAP
Foresight Associate Editor
Author of Demand Forecasting for Managers with Enno Siemsen

Will Deep and Machine Learning Solve Our Forecasting Problems?

Deep Learning and other approaches in Machine Learning are having a tremendous impact on our life, from voice assistants to all kinds of image recognition tasks. Similarly, there have been more and more success stories of ML applied to forecasting tasks. Since there is no shortage of loudly proclaimed enthusiasm, I will draw attention to some aspects where DL/ML does not (yet) solve all our problems, as well as point out some issues where the ML hype actually helps “classical” forecasters learn new tricks. I anticipate vocal counterarguments and a lively, entertaining discussion.

Spyros Makridakis
Creator of the M-Competitions

Human Intelligence (HI) Versus Artificial Intelligence (AI)  and Intelligence Augmentation (IA)

HI and AI, in its present form, are two dissimilar but complimentary forms of intelligence, with AI surpassing HI in games and images and approaching it in speech and text recognition while HI excels in most other tasks that involve situations/decisions where the rules are not known and the environment can change. The two critical questions are if AI and HI can coexist each contributing where they excel while avoiding their weaknesses? IA presents the possibility of substantial improvements in HI by exploiting progress in AI and advances in the related technologies of nanotechnology and neuroscience.

Nada Sanders
Distinguished Professor, Supply Chain Management
Northeastern University

Humachine: The Enterprise of the Future

The Humachine represents the optimal human-machine partnership. Here, people and technology operate seamlessly. This hybrid workforce creates an unparalleled team to create industry leaders. It is an enterprise that takes advantage of technology to complement and augment human decisions. “Machines”– technologies such as AI – are stronger, better, faster, more precise. Humans are intuitive, creative, and understand context. Harnessing the strengths of both – and finding the right way for them to work together – is the recipe for success. This talk is about how an enterprise creates and manages synergies between technology and people to optimize decision-making, innovation, and performance, the roadmap for getting there, and what individuals can do to prepare themselves for work in the enterprise of the future. The future is already here.

Shaun Snapp
Author of Supply Chain Systems

The Data Implications for AI/ML Projects

AI/ML brings up many data-related issues that tend to be glossed over. This presentation will address cover this neglected area. Among the questions we’ll discuss are:

1. Do we need specialized databases to support AI?
2. How much do Big Data and Data Lakes rely on the benefits of AI?
3. What are the issues with data availability and data development timelines for AI projects?
4. Are the data requirements for AI difference from those for statistical forecasting projects?

Eric Stellwagen
President, Business Forecast Systems, Inc.

A Winning Combination for Increasing Accuracy: AI-based Automatic Forecasting & Domain Knowledge

Many companies rely upon automatically-generated forecasts to drive their demand planning processes. But the quality of the automatic forecasts varies dramatically depending upon the approaches built into the software that these organizations use. Eric will survey how automatic forecasting algorithms typically work—pointing out their strengths and weaknesses—and review the key elements needed to develop a successful automatic approach. He will show how you can improve accuracy by combining strong AI-based automatic forecasting approaches with your domain knowledge. You will leave the session not only with a better understanding of automatic approaches, but with insight about how to evaluate software solutions offering automatic forecasting and pragmatic tips for getting the most out of the software you currently use.

Rob Stevens

VP and Principal, First Analytics

Machine Augmented Demand Planning

Setting hype around machine learning and AI aside, these technologies are now finding success in assisting humans with forecasting and demand planning. We present a framework which classifies various applications for AI and closely related analytical tools. The framework is comprised of three broad application areas: data engineering; forecasting models; and planner decisions. Within each of these areas are sub-categories, and we lay out more than a dozen application cases. For a selection of these cases, we illustrate with real-world examples.

Larry Vanston
President, Technology Futures

Forecasting Artificial Intelligence

Artificial Intelligence will likely revolutionize our world, yet published forecasts on AI performance, adoption, and impacts are surprisingly weak given the high stakes. Here we report on TFI’’s work to correct this, including our current forecasts and work in progress. This includes quantitative forecasts of AI performance improvement and an analysis of the drivers and constraints for further AI adoption. We will also address the impacts and implications of AI for employment and our future.  Finally, we discuss AI’s role in the future of forecasting, given the recent significant progress in that area.

Bernie Wang
Senior Machine Learning Scientist
Amazon AI Labs

Neural Forecasting at Amazon: Models, Tools and Applications

Time series forecasting is a key ingredient in the automation and optimization of business processes: in retail, deciding which products to order and where to store them depends on the forecasts of future demand in different regions.  In cloud computing, the estimated future usage of services and infrastructure components guides capacity planning.

Recent years have witnessed a paradigm shift in forecasting techniques and applications, from computer-assisted assumption-based models to data-driven and fully-automated approaches. This shift can be attributed to the availability of large, rich, and diverse time series data sources but the result is a new set of challenges. In this talk, we give an overview of state-of-the-art Neural Network forecasting models, covering the underlying principles, their usages in various forecasting domains, and corresponding tools (such as GluonTS, DeepAR in Amazon SageMaker, Amazon Forecast, etc.). We shall also discuss recent advances of neural network forecasting, in particular, the models that efficiently combine the expressive power of neural networks and the data efficiency of traditional probabilistic models.