Join this live course to master ML strategy with Foster Provost
Learn how to craft ML products and services that increase customer engagement and resource efficiency. Live with Professor Foster Provost and a community of likeminded product and engineering leaders.
5 x 2 hr sessions (+ recordings of each live session)
4pm PDT; May 3, 5, 11, 12, 17
Building ML products and services can significantly increase customer engagement and resource efficiency. Realizing this impact requires ML strategy that extends well beyond the technical implementation
As part of Foster's cohort you will master frameworks that allow you to:
Identify which problems in your business can (and can’t) be solved using ML technologies
Evaluate the ROI of ML-based projects (rather than the traditional academic-style evaluations)
Building ML products and services can significantly increase customer engagement and resource efficiency. However, realizing this impact is a major challenge. Throughout this course we will develop and apply several frameworks for reliably generating ROI from ML-based projects.
Identify which problems can (and can’t) be solved using ML technologies
Manage the high degree of uncertainty in ML projects
Product Managers who want to create new products/services based on machine learning or incorporate machine learning into existing products.
Entrepreneurs and investors who need to understand whether and how ML can create business value from data assets.
Senior Data Scientists and Engineering Managers who need to understand how to use ML to realize value from their data assets, or how to invest in data assets to increase business value.
Foster Provost is Ira Rennert Professor of Entrepreneurship and Data Science at NYU and a long-time ML practitioner. He is also a Distinguished Scientist at real-estate tech giant Compass. He co-founded a half-dozen successful ML-based startups, conceiving and orchestrating their foundational ML architectures and methods—most notably data science powerhouse Dstillery and recent IPO Integral Ad Science (IAS). His innovation-in-practice also has won awards—for example, significant awards for applied ML work with Verizon, AT&T, Spotify, and Citibank.
Professor Provost has been developing leading curricula for decades. Most notably he developed pioneering MBA data science classes at NYU’s Stern School, as well as the intro classes for NYU’s MS in Business Analytics and MS in Data Science degrees. His book Data Science for Business is the gold standard for serious examination of realizing business value with advanced statistical methods and was listed as one of Fortune Magazine's "must read books for MBAs."
Professor Provost is also one of the world’s leading ML scholars. His research is widely used and highly cited and has won numerous awards, including the ACM SIGKDD Test of Time Award, the INFORMS Design Science Award, and best paper awards in top venues across four decades.
Realizing business impact from ML is a major challenge. Top firms can have spectacular failures — think Watson’s debacle with MD Anderson, or Zillow’s iBuyer collapse. On the other hand, the best ML professionals can consistently get business impact from ML. What gives?
Getting business value from ML is a craft, and good craftspeople have good tools. In this cohort we will develop a half-dozen conceptual tools — frameworks to help us think data-analytically and thereby greatly increase our likelihood of succeeding with ML. Using these tools, we will examine a variety of real applications and work through case studies. The tools help us to answer the key questions of creating ML products and services:
Session 1 - Mapping the problem space
In this introductory lecture we develop three conceptual tools that help us map any problem space and understand where ML could produce impact:
Laying out the high-level goals of a business and it's data foundation
Understanding the key distinction between ML and AI inference with ML models
Decomposing and organizing the process of creating ML solutions
Session 2 - Measuring success
In this lecture we discuss real examples of ML solutions and the limitations of traditional academic style reporting. Specifically we will:
Walk through a detailed case study analyzing the metrics that ML practitioners report, their limitations for actual ML products, and how to connect them to product needs and outcomes
Discuss the important ML concepts of overfitting, complexity control, and the bias/variance tradeoff
Examine a case study of creating a new ML cross-sell product for one of the largest banks
Session 3 - Preparing for unique challenges
In this lecture we will go into the challenges of managing ML products and why they are different from the products that many of us are familiar with. Specifically we will:
Explore how to build customer trust in ML-based products (with example cases from IBM, Dstillery, and Compass)
Explore privacy and ethical implications of ML-based products
Examine when ML products go wrong (using Zillow’s iBuyer as a case study)
Session 4 - Re-evaluating the problem formulation
In this lecture we build upon our simple problem formulation and ML solution framework that we developed in Sessions 1 and 2. Specifically we will:
Introduce the Expected Value framework that can reveal how to integrate many pesky complications of real-world ML applications
Discuss causal decision making (with a case study of using ML to generate playlists at Spotify)
Session 5 - Strategy & Management Frameworks
In this lecture we explore: how do we choose where to apply our limited ML resources? Which are most likely avenues to success? Specifically we will:
Introduce a framework for estimating the expected ROI for ML products
How to work through typical ML strategy questions - choosing between potential products, dealing with uncertainty, how to create competitive advantage (with a case study from Integral Ad Science)
Explore how to build and retain effective ML teams
You don’t! We record every live session in the cohort and make each recording and the session slides available on our portal for you to access anytime.
Each learner receives a certificate of completion, which is sent to you upon completion of the cohort (along with access to our Alumni portal!). Additionally, ScholarSite is listed as a school on LinkedIn so you can display your certificate in the Education section of your profile.!
Throughout the cohort, there may be take-home questions that pertain to subsequent sessions. These are optional, but allow you to engage more with the instructor and other cohort members!
While we cannot guarantee that your company will cover the cost of the cohort, we are accredited by the Continuing Professional Development (CPD) Standards Office, meaning many of our learners are able to expense the course via their company or team’s L&D budget. We even provide an email template you can use to request approval.