Human-AI Co-Mentorship in Project-Based Learning: A Case Study in Financial Forecasting

Problem
This paper addresses the gap in literature regarding the integration of AI tools in project-based learning environments, particularly for high school and early undergraduate students with limited backgrounds in AI and finance. The authors present a case study that explores how AI can facilitate mentorship and enhance learning outcomes in a financial forecasting project. This work is a preprint and has not undergone peer review.

Method
The core technical contribution of this study lies in its pedagogical framework, which emphasizes workflow design over traditional classroom instruction. The students engaged in a project focused on ETF price prediction, utilizing AI tools to assist in coding and problem-solving. The methodology involved iterative interactions with AI systems, allowing students to develop code collaboratively while receiving real-time feedback during daily stand-up meetings. This approach enabled students to define problems and explore solutions in a hands-on manner, fostering deeper engagement with both computer science and finance concepts.

Results
The authors report that all participating students successfully developed functional models for financial forecasting, demonstrating significant progress in their understanding of both AI and market analysis. While specific quantitative metrics are not provided, the qualitative outcomes indicate that students were able to leverage AI tools effectively, resulting in meaningful advancements in their projects. The study highlights the potential for AI to enhance learning experiences, particularly in complex domains like finance, where traditional educational methods may fall short.

Limitations
The authors acknowledge several limitations, including the small sample size of participants and the lack of a control group to compare the effectiveness of AI-assisted learning against traditional methods. Additionally, the study does not provide detailed performance metrics or benchmarks to quantify the improvements achieved through AI tools. An obvious limitation not discussed by the authors is the potential variability in student engagement and prior knowledge, which could affect the generalizability of the findings.

Why it matters
This research has significant implications for the design of educational programs that incorporate AI tools, particularly in STEM fields. By demonstrating that high school students can effectively engage with complex topics through AI-assisted mentorship, the study suggests a paradigm shift in how educational institutions might approach project-based learning. The findings advocate for the integration of AI in curricula to enhance student autonomy and problem-solving skills, potentially leading to more innovative and effective learning environments. This work could inform future research on AI’s role in education and its application in various domains.

Authors: Freyaa Chawla, Ahan Chawla, Rishi Singh, Joe Germino, Grigorii Khvatskii
Source: arXiv:2605.05144
URL: https://arxiv.org/abs/2605.05144v1