Sophomore Hansheng Zhu formed the AI@Penn Club to connect academia to industry.
To accomplish the above, Zhu and the team created four branches: development, research, outreach, and incubation.
“We’re building products to work with industry and startups,” Zhu said of the development branch. One such startup includes a global talent management platform. While Zhu could not disclose the name, he shared salient details on how it will support both applicants and employers. From the applicant side, AI agents will conduct mock interviews, responding with tailored suggestions “to guide them through their potential career paths.” The AI agents will then allow employers to determine if applicants fit before scheduling interviews.
Still a work in progress, the team hopes to optimize the talent management platform for March 2025 completion, Zhu explained.
The AI@Penn Club’s research branch works on open-source collaborations with Penn Engineering alumni in industries such as Apple and Google. According to Zhu, Apple recently released their version of a text-to-image generation model. This provides a very foundational functionality for Siri. As such, the release provided a “complete framework for generating image to text. Before that, there wasn’t really a stable diffusion model that could be applied to any Apple or silicon device,” Zhu said, adding, “it’s still under a lot of testing.”
As the name suggests, the AI@Penn’s outreach branch entails hosting both social and informational events. “This will give participants the opportunity to network and make industry connections.”
And when it comes to incubation, the AI@Penn Club wants to help facilitate the path from research and development to industry through Penn resources such as Venture Labs.
In many ways, the goals Zhu holds for the AI@Penn Club mirror his professional goals.
“Currently, my dream company would be Apple,” Zhu said. “They are not leading the way in AI because they started a bit slow [but] they’re making inroads.” Zhu’s comments refer to their generative image diffusion models utilized for Siri. “My idea is to adapt their models to smaller versions so they could be used on our iOS devices.”
In that regard, Zhu would like the accomplishments of Penn’s Engineering school to become analogous to the accolades associated with elite West Coast universities. According to Zhu, their students have more access to opportunities given their proximity to Silicon Valley.
From Zhu’s perspective, the AI degree offered at Penn, like AI overall, has boundless opportunities. Take the recent sharing of DeepSeek, for example. As Zhu explained, DeepSeek’s open-source AI has optimized the computational efficiency of LLMs, reducing inference latency while maintaining state-of-the-art performance on benchmark datasets. This advancement not only enhances real-time applications but also lowers operational costs, making cutting-edge AI more accessible to researchers and developers globally, Zhu said.
From a personal perspective, Zhu’s pivot from the CIS to AI undergraduate degree at Penn speaks directly to his academic interests. “I grew up in Hong Kong and went to an international high school,” Zhu explained, adding “my background was in mathematics and natural science.” That background in mathematics, and subsequent awards, paved the way for his acceptance into Penn’s CIS program. Though the skills associated with the AI program in machine learning and natural language processing proved a better fit for Zhu. “The math behind algorithms for different AI tasks—that’s what fascinates me.”
To date, the AI Club has received over 100 applications. “We would like it to grow to 200,” Zhu said. To that end, it hosted an information session which garnered significant interest. Each applicant applies to the branch of AI that speaks to their skills. Based on applicant response, AI Club leadership is hosting its first round of interviews.
To learn more about the AI Club at Penn: ai-at-penn-main-105.vercel.app