Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Empa: An AI-Powered Virtual Mentor for Developing Global Collaboration Skills in HPC Education

Ashish, Aparajita Jaiswal, Sudip Vhaduri, Niveditha Nerella, Shubham Jha

TL;DR

The paper addresses the need for intercultural collaboration training in HPC education. It introduces Empa, an AI-powered virtual mentor that uses a lightweight LLM fine-tuned with $4$-bit QLoRA on a $3$B-Instruct model to provide personalized, culturally aware feedback within six modular activities. The authors combine Hofstede's cultural dimensions and Bennett's DMIS with portable intercultural modules to deliver scalable, discipline-agnostic training for computing students. They demonstrate feasibility and provide guidance for curricular integration and pilot deployment, highlighting potential for broad adoption in technical education to produce globally competent HPC teams.

Abstract

High-performance computing (HPC) and parallel computing increasingly rely on global collaboration among diverse teams, yet traditional computing curricula inadequately prepare students for cross-cultural teamwork essential in modern computational research environments. This paper presents Empa, an AI-powered virtual mentor that integrates intercultural collaboration training into undergraduate computing education. Built using large language models and deployed through a progressive web application, Empa guides students through structured activities covering cultural dimensions, communication styles, and conflict resolution that are critical for effective multicultural teamwork. Our system addresses the growing need for culturally competent HPC professionals by helping computing students develop skills to collaborate effectively in international research teams, contribute to global computational projects, and navigate the cultural complexities inherent in distributed computing environments. Pilot preparation for deployment in computing courses demonstrates the feasibility of AI-mediated intercultural training and provides insights into scalable approaches for developing intercultural collaboration skills essential for HPC workforce development.

Empa: An AI-Powered Virtual Mentor for Developing Global Collaboration Skills in HPC Education

TL;DR

The paper addresses the need for intercultural collaboration training in HPC education. It introduces Empa, an AI-powered virtual mentor that uses a lightweight LLM fine-tuned with -bit QLoRA on a B-Instruct model to provide personalized, culturally aware feedback within six modular activities. The authors combine Hofstede's cultural dimensions and Bennett's DMIS with portable intercultural modules to deliver scalable, discipline-agnostic training for computing students. They demonstrate feasibility and provide guidance for curricular integration and pilot deployment, highlighting potential for broad adoption in technical education to produce globally competent HPC teams.

Abstract

High-performance computing (HPC) and parallel computing increasingly rely on global collaboration among diverse teams, yet traditional computing curricula inadequately prepare students for cross-cultural teamwork essential in modern computational research environments. This paper presents Empa, an AI-powered virtual mentor that integrates intercultural collaboration training into undergraduate computing education. Built using large language models and deployed through a progressive web application, Empa guides students through structured activities covering cultural dimensions, communication styles, and conflict resolution that are critical for effective multicultural teamwork. Our system addresses the growing need for culturally competent HPC professionals by helping computing students develop skills to collaborate effectively in international research teams, contribute to global computational projects, and navigate the cultural complexities inherent in distributed computing environments. Pilot preparation for deployment in computing courses demonstrates the feasibility of AI-mediated intercultural training and provides insights into scalable approaches for developing intercultural collaboration skills essential for HPC workforce development.

Paper Structure

This paper contains 18 sections, 4 figures.

Table of Contents

  1. Introduction
  2. Technical Research Origins
  3. Educational Adaptation and Use Case
  4. System Architecture Overview
  5. Empa LLM Architecture
  6. Empa User Interface
  7. Backend Architecture and Implementation
  8. API Design and Endpoints
  9. Data Persistence Layer
  10. LLM Integration Architecture
  11. Security and Scalability Considerations
  12. Error Handling and Resilience
  13. Performance Optimization
  14. Module Content and Intended Outcomes
  15. Curricular Integration Strategy
  16. ...and 3 more sections

Figures (4)

  • Figure 1: Training loss curve showing convergence of the fine-tuned LLaMA 3 3B model during QLoRA fine-tuning.
  • Figure 2: Onboarding screen where learners enter their academic details to personalize the Empa experience.
  • Figure 3: Empa chatbot greeting user with a personalized message and overview of the collaboration journey.
  • Figure 4: Drag-and-drop activity for classifying characters by communication style, reinforcing key cultural concepts.