HBS Foundry will be a digital platform that trains and coaches underserved entrepreneurs, including underrepresented minorities, women, and founders in emerging markets or U.S. cities that aren’t startup hubs. Foundry will feature: 1) modular content – video rich and practical – providing comprehensive coverage of the concepts and skills that an entrepreneur needs to master; and 2) AI-powered applications that will guide an entrepreneur to relevant content, serve as a tutor, and advise the entrepreneur on key venture decisions. HBS’s goal with Foundry is to dramatically increase the School’s impact on entrepreneurs, helping to provide opportunity for all and showcase business as a force for good.

In Foundry AI Lab the instructors — the faculty co-leads for HBS Foundry — will work with small teams of students who will develop, test, and refine prototypes of AI applications that will eventually be built into Foundry. Through learning-by-doing, students will gain skills with applying GenAI technologies and deeper understanding of the capabilities and limitations of these important new technologies. Students will also gain familiarity with challenges confronting underserved entrepreneurs. Since most MBAs will work extensively with GenAI technologies in the future, the course should be of interest to a broad range of students. It may be of special interest to those who plan to work as product managers, in entrepreneurial ventures, or for organizations that aim for social impact.

Some examples of applications we expect to prototype in Foundry AI Lab include:

• Bots that play the counterparty role in different entrepreneurial negotiation settings, e.g., splitting equity with a co-founder; negotiating terms with an investor; hiring a key early employee. The bots could also provide feedback on the learner’s negotiation approach.
• A bot that assumes the role of a targeted persona for customer discovery interviews, providing insight on unmet customer needs and feedback on the learner’s questioning approach.
• An avatar that answers a learner’s question from the perspective of a specific domain expert – say, an HBS professor fielding questions about startup sales.
• A bot that provides feedback on an entrepreneur’s pitch deck and suggestions for improvement.

In pairs, students will identify and evaluate the effectiveness any existing 3^rd-party AI-powered applications that target the entrepreneurial tasks they’ve been assigned. Teams will then prototype their own bot or tool and aim to improve its reliability and functionality through iterative prompt engineering; by experimenting with different source content and instructions provided to Retrieval-Augmented Generation-powered Large Language Models; and through user testing.

The first half of our weekly two-hour class sessions will feature “how to build with GenAI” talks by guest speakers, including staff from HBS Digital Transformation Team and the Digital Data Design Institute at Harvard. These speakers and other experts will be available to coach student teams. The second half of our class sessions will be devoted to providing peer feedback on teams’ work-in-progress.

During the semester, teams will submit a workplan and periodic updates on their work-in-progress. Their final deliverable will be a prototype demonstrated during a class presentation, accompanied by a report evaluating the prototype’s strengths and limitations and detailing recommended next steps for the application.

Copyright © 2024 President & Fellows of Harvard College. All Rights Reserved.