Essa Imhmed
Tenure Track Assistant Professor | Researcher
Projects
From Submission to Feedback: A CI-Driven Automated Code Assessment
Timely, formative feedback is essential in academia, helping students reinforce learning and develop the skills needed for subsequent assignments and advanced topics. However, manually assessing code submissions is time-consuming and often results in delayed feedback. This project investigates state-of-the-art automated code assessment techniques and proposes CodeInspector—a CI-driven, semi-automated solution that leverages industry-standard tools to assess student code submissions and generate immediate feedback on quality and correctness, while supporting iterative learning and continuous improvement.
Status: Funded by the ENMU FRID Grant Program (2023-2025); Amount: $8471; Role: Principal Investigator.
Related Publications
- Essa Imhmed, Edgar Ceh-Varela, Ludwig Scherer, George Candal, and Ivan Sanjaya. WIP: CodeInspector: automated LLM-Supported CS1- Level code assessment. In 2025 IEEE Frontiers in Education Conference (FIE) (FIE 2025), page 4.98, Nashville, USA, November 2025.
Students
- Ludwig Scherer (2025 - current) – Undergraduate Research Assistant on the CodeInspector project.
- Scott Kilgore (2023 - 2024) – Undergraduate Research Assistant on the CodeInspector project.
AI-Driven Static Analysis and Feedback Generation
A common challenge among novice programmers is interpreting runtime error stack traces and manually detecting logic and style errors in their Java code, often leading to inefficient debugging and poor coding practices. This study aims to design an AI-driven static analysis and feedback-generation approach that leverages natural language processing (NLP) and machine learning clustering techniques to statically identify code errors.
Status: Funded by the ENMU FRID Grant Program (2025–2026); Amount: $2,046; Role: Principal Investigator.
Students
- George Candal (2025 - current) – Undergraduate Research Assistant on the project.
- Ivan Sanjaya (2025 - current) – Undergraduate Research Assistant on the project.
Local Memory Store for Embedded Systems
Local Memory Store (LMStore) is a novel scratchpad memory (SPM) design, featuring some hardware management of the SPM along with explicit access and interaction by the program. Rather than presenting the program with a flat, directly accessed address space, LMStore uses indirect references in the form of 2-tuples and 3-tuples to access data stored in its memory, allowing LMStore to manage its memory more effectively for better utilization. This project explores data management schemes specifically designed to maximize LMStore performance, with a focus on reducing power consumption and minimizing memory access latency in embedded programs.
Related Publications
- Imhmed, E., Ceh-Varela, E., Cook, J., & Parten, C. (2023, June). Evaluation of the performance impact of SPM allocation on a novel scratchpad memory. In 2023 IEEE 47th Annual Computers, Software, and Applications Conference (COMPSAC) (pp. 972-973). IEEE.
- Imhmed, E., Cook, J., & Badawy, A. H. (2022, September). Evaluation of a novel scratchpad memory through compiler supported simulation. In 2022 IEEE High Performance Extreme Computing Conference (HPEC) (pp. 1-7). IEEE.
Status: Completed; interested in more.
Students
- Caleb Parten (Spring 2023) – Undergraduate Research Assistant on the LMStore project.