In the May 15 Women in Ag Tech webinar, Dr. Susan Olsen, Founder of Action Intel, shared how generative AI tools like Claude.ai and ChatGPT can support predictive analytics, even for those without a formal data science background.
Dr. Olsen highlighted Claude’s advanced capabilities in summarizing structured documents such as MSDS (Material Safety Data Sheets), but she also emphasized its usefulness—and that of ChatGPT—in creating predictive models using datasets such as USDA grain train speeds. Depending on the task, both tools can write Python code for classifier models like random forests or XGBoost, helping users forecast outcomes such as rail speed by month or operator.
“ChatGPT will write your Python code to do that,” Dr. Olsen explained. “You can even guide it to use specific variables like operator and month of year.”
While working with smaller datasets due to free-tier limitations, she demonstrated how to upload samples and prompt the AI to build models. Depending on the AI model, results might come in different coding languages—like Python or React—but users can always request the format they need.
This hands-on example shows how generative AI is evolving from content generator to data science assistant, opening doors for more women to explore predictive modeling across ag tech and beyond.
