
Hosted by Robert Clements · EN

In this episode, Rebekah and Vitoria share how the Bootcamp EDA project pushed them beyond “just making plots” and into building a clear, credible narrative that a real stakeholder could act on. They talk through how they chose what to show, how they structured the story, and how design and interpretation choices can change what an audience takes away. If you’ve done EDA before but want to communicate insights with more impact, this one is for you.

Cody and Robert speak with Shan Wang, the Program Director for the Master of Science in Data Science program at University of San Francisco, about data science education and where it may be going during these quickly changing times.

In the final episode of Cohort 12's student series, we talk with Ian Duke and Tatshini Ganesan about their year in the MSDS program. Ian discusses his practicum at the ACLU training ML models on massive amounts of police body cam footage to flag videos for specific topics like searches or arrests. Tatshini shares about her experience working with Vibrant Data Labs to disseminate climate change-related financial data to companies and researchers who need it. They also give their insights on working in hybrid environments, navigating the work/life balance, and forming community and friendships in MSDS.

In this episode we talk to two more students from our most recent cohort, Bassim and Rishi, about their experiences in the MSDS program. Bassim shares some insights into how he has optimized his job search process with some noticeable success, and Rishi walks us through his data engineering practicum and decision to enroll in the data engineering concentration.

In this episode we talk to two of our graduating MSDS students, Amy and Rashmi, about their experiences in the MSDS program, their strategies for selecting their practicum companies, and the things that most surprised them about getting their MS degrees in Data Science at USF.

The top three winners of our Advanced ML course Kaggle competition join us to share their strategies and lessons learned from training models on real estate listing price data from houses in Spain. Hear how they navigated the decisions involved in model building, including feature engineering, the curse of dimensionality, and the challenge of handling text descriptions in a foreign language.

In this episode, we chat with Kekona Sorenson, a versatile data science leader at Microsoft. Kekona shares insights on his team’s role in improving Microsoft products through data analysis and experimentation. We also discuss the importance of cultural shifts in data-driven decision making and the need for effective communication in Data Science. According to Kekona, "the thing that separates an average Data Scientist from a great Data Scientist is your soft skills". He also highlights the importance of building trust with stakeholders, and he attempts to sell the role of Product Data Science to an audience enraptured by all things Gen AI.

In this episode, we speak to Nico Thiébaut, a seasoned machine learning engineer with a PhD in Physics. Nico shares his experiences working on the popular gaming platform Roblox, highlighting the platform's growth and the challenges of creating engaging, user-generated content. We also discuss the role of machine learning in game development and the ethical complexities of content moderation on large platforms. Our hosts, Cody Carroll and Robert Clements add their insights on human moderation and the importance of soft skills in consulting.

In this episode, we speak to Sundar Dorai-Raj, a seasoned data scientist at Google and MSDS faculty at USF. He shares about his experience working on Bard/Gemini and the constraints and freedoms of working at a major tech company. We also have a discussion about the place of LLMs and Generative AI in academia — for both students and professors.

In this episode, MSDS faculty Mustafa Hajij gives a beginner-friendly introduction to topological deep learning. We discuss the foundations and some applications of his research, including graphs, social media networks, and chemical interactions of drugs. He also shares his take on some new developments in Generative AI and discusses how he incorporates emerging topics in deep learning into the classes he teaches at USF.