Welcome to the Agentic and Dependable AI Lab (ADAL), an undergraduate research lab in the Department of Computer Science at Elizabethtown College. Students in ADAL build agentic AI systems and study how to make AI dependable: secure, private and reliable, from edge devices to the cloud. Our current research focuses on four areas:
ADAL students have co-authored papers at IEEE and ACM venues, presented at the ACM SIGCSE Student Research Competition, and gone on to Ph.D. programs and industry careers. The lab’s research has been supported by Intel, Google, the National Science Foundation, and Elizabethtown College’s SCARP and Faculty Grant programs.
The lab, formerly the Advanced Data Analytics Lab, is led by Peilong Li, Ph.D., Associate Professor of Computer Science and Program Director of the Computer Science and Data Science Programs. Learn more on his personal website.
What we look for in student researchers today (you don’t need all of it to start)
Builds with Claude Code, Codex or similar agents: plan the work, delegate it, iterate
Treats agent output as a draft: reads it, tests it, and owns the result
Specs, Git branches, code review, tests and CI across a large codebase
In any language: you can read, debug and reason about code you didn’t type
Our research areas: machine learning, LLMs, security and edge systems
Writes clear specs, papers and talks; discloses AI use and protects sensitive data
Student research internships to prototype a self-hosted large language model system for the college (2026-2028).
Vertical federated learning to predict student success without sharing raw records across campus offices (SCARP 2026).
AI-powered, real-time interactive exercise support for elderly communities on low-power edge devices (SCARP 2025).
Proficiency-adaptive Japanese learning through NLP-powered translation and a browser extension.
An open-source, coprocessor-agnostic vision pipeline for the FIRST Robotics Competition (SCARP 2023).
Edge-network malware detection with deep learning for smart homes and communities (Google Cloud Research Award 2022-2023).
Machine-learning pipelines for detecting malware, supported by SCARP 2022 and an Intel Academic Grant.
Data science for predictive diagnosis in industrial IoT (IIoT)-enabled smart manufacturing, with industry mentors (SCARP 2021).
A performance study of machine learning and deep learning models for detecting encrypted malware traffic (SCARP 2020).
An AI-driven crawler rendered in 3D that self-teaches itself to collect rewards.
Classify encrypted network traffic with accelerated machine learning models.
Publications by ADAL -ers