When OpenAI released ChatGPT, Ghazaleh Sadooghi, EWMBA 25, wanted to use it but feared being labeled a cheater.
A senior software engineer at LinkedIn, she knew there had to be a better way to incorporate AI ethically into academia. So she and her husband, Mo Zadeh, MIMS 16 (information systems), began developing the startup Rumi Technologies, aiming to empower students to use AI responsibly within guidelines set by their instructors.
In this interview, Sadooghi explains how Rumi enables responsible AI use and discusses how Rumi is already being used by Haas students.
Tell us a bit more about what Rumi does.
Rumi functions like the suggested changes feature in Google Docs. We included a time slider that reveals every edit and AI prompt used during the writing process. This is all seamlessly integrated into Canvas, a new interface for working with ChatGPT on writing and coding projects. So professors can use Rumi to customize the AI prompts and functionalities for their assignments. Students who write their papers within the Rumi platform can show how they used tools like ChatGPT, Claude, and Gemini, in their research. Instead of relying on flawed AI detectors or taking a hands-off approach to AI, professors get insights into students’ writing and thinking processes, rather than just evaluating final submissions.
Which Berkeley Haas programs are trialing Rumi?
Berkeley Haas is using Rumi in undergraduate and MBA classes. The feedback we’ve received from students and faculty has been very helpful in guiding our product development. We have the support from the Haas Digital team, particularly Tom Tripp, executive director of Haas Digital, and learning support consultant Hamza Taha. In addition to Haas, we’ve also signed 10 contracts with leading educational institutions in the United States, including Tufts University, Foothill College (which Niche ranked the No. 1 community college in California), and The Baldwin School (the No. 2 private high school in Pennsylvania).
What are some key challenges schools face when incorporating AI?
We’ve engaged with over 1,000 educators and instructional designers across the United States and internationally. In most schools, we work with their AI steering committees that have formed to gather faculty feedback and propose solutions on how to deal with AI. Within each committee, there is often a divide: Some members advocate for banning AI to uphold academic integrity, while others support allowing AI to promote AI literacy.
Should AI policies be set by institutions?
We believe AI policy shouldn’t be set at the institutional level. It should be determined at the assignment level at the discretion of individual instructors. Schools should provide broad guidelines on AI policy and offer tools to implement these policies for any assignment. Rumi supports a range of AI policies, from no AI usage to partial AI integration to full AI integration.
How did your time at Berkeley Haas help you launch your startup?
When we had the initial idea for Rumi and a rough prototype, I shared it with one of my favorite professors, Maximilian Auffhammer. He was excited about our vision and provided valuable feedback, becoming a source of motivation for our team. Later, when we faced challenges in defining our pricing strategy, I turned to professional faculty member Bill Pearce, who guided us to think beyond simply making the product cheaper to appeal to customers.
The Berkeley Haas Entrepreneurship Program (BHEP) and the UC LAUNCH accelerator program were crucial to our startup’s progress. They connected us with top mentors, provided valuable exposure, helped us secure grants and small angel investments, and introduced us to our very first outside investors, Underdog Labs, who are very supportive. The team at LAUNCH, especially BHEP executive director Rhonda Shrader, our mentor, Jed Katz, MBA 96, and Mohammad Naqvi, MBA 24, were incredibly helpful and supportive. To date, we’ve raised $450,000 finished first in UC Launch’s 2023 startup cohort, and have been featured in Forbes.
What’s the long-term vision for Rumi?
Our goal is to build the world’s leading academic integrity and artificial intelligence (AIAI) platform for educational institutions. We already have interest from schools in the Philippines, Thailand, and Australia. Beyond EdTech, we see significant opportunities in other industries, such as workforce development and legal tech.
Posted in:
Topics: