Software Engineer PhD Student · Idaho State University

Costain Nachuma

I ship production developer tooling — IDE plugins, LLM-integrated pipelines, graph-grounded retrieval systems — and study how AI reshapes software engineering at scale. 9+ years of prior industry experience; currently researching AI-assisted code comprehension and supply chain security.

6
Publications

4.0
GPA

9+
Years Industry

500+
Users Supported

4
Awards Won

About

🛠️
Production AI Tooling
Shipped a 5-layer Eclipse IDE plugin spanning three LLM backends, released publicly with a Zenodo artifact
📊
Large-Scale Systems
Python/DuckDB and Neo4j pipelines processing 14M+ records — GitHub, Maven, and Chromium ecosystems
🔒
Supply Chain Security
Quantifying dependency risks and evaluating SBOM tools across open-source ecosystems

Software engineer and PhD student in Computer Science at Idaho State University (GPA: 4.0) who ships production tooling end-to-end: designed and released Comprendia, a 5-layer Eclipse IDE plugin integrating three LLM backends, and built Python/graph-database pipelines processing 14M+ records. Before academia, I spent 9+ years leading engineering and infrastructure for a 500+ user government network at the National Assembly of Zambia. My research on AI-assisted software engineering is published at top venues including ICSME 2026, MSR 2026, MSR 2025, MSR 2024, and ACM SAC 2024. I bring both rigorous research depth and real production experience.

Selected Projects

Comprendia — AI-Augmented Code Comprehension
Eclipse IDE plugin · Java · 5-layer architecture

Designed and shipped a 5-layer IDE plugin combining a Cytoscape.js dependency-graph UI, a JDT AST analysis engine, OSV.dev/JGit vulnerability integration, and an AiGateway spanning three production LLM backends (Claude, OpenAI, Ollama). Introduces Graph-Aware Callee Pruning (GACP), an algorithm that grounds LLM prompts in a live dependency graph instead of embedding similarity — producing deterministic, auditable output reproducible across LLM families via an O(V+E) reachability filter. Benchmarked across 6 code fixtures × 4 prompt modes × 3 LLM families. Tool demo accepted at ICSME 2026.

JavaEclipse JDTCytoscape.js Neo4jLLM IntegrationOSV.dev
Chromium Bug ETL Pipeline
Python · 5-script scraping & parsing pipeline

Built a 5-script scraping/HTML-parsing pipeline against the live Chromium issue tracker, working around its 1,000-record API cap to assemble a complete research dataset. Powered a study of 5,575 security and 4,605 performance bug reports, published in "On the Assessment of Security and Performance Bugs in Chromium Open-Source Project."

PythonBeautifulSoupREST APIsData Engineering

Professional Experience

Graduate Research & Teaching Assistant
Idaho State University — Pocatello, ID
Aug 2023 – Present
  • Designed and shipped Comprendia, an Eclipse IDE plugin with a 5-layer architecture (Cytoscape.js UI, JDT AST analysis engine, OSV.dev/JGit integration, and an AiGateway spanning three production LLM backends — Claude, OpenAI, Ollama); benchmarked across 6 code fixtures × 4 prompt modes × 3 LLM families and released publicly with a Zenodo artifact and demo video.
  • Built Graph-Aware Callee Pruning (GACP), an algorithm that grounds LLM prompts in a live dependency graph instead of embedding similarity, producing deterministic, auditable output reproducible across LLM families (O(V+E) reachability filter).
  • Built a Python/DuckDB pipeline processing 33,596 GitHub pull requests across 2,807 repositories, and a separate Neo4j-backed pipeline querying 14.4M+ package releases for dependency-vulnerability analysis (21-hour batch job on a 64GB/24-core system); mentored 20+ students.
  • Conducted a Systematic Mapping Study (Kitchenham's guidelines) synthesizing literature on software visualization tools.
Instructor of Record — CS 3309: Advanced Object-Oriented Programming
Idaho State University — Pocatello, ID
Fall 2025
  • Sole instructor for a 45-student undergraduate course; independently designed the curriculum, delivered all lectures, and graded consistently — rated 3.50/4.0 for overall teaching effectiveness (90% agree/strongly agree) in end-of-term evaluations.
Senior Support Specialist, Software Engineer
National Assembly of Zambia — Lusaka, Zambia
Aug 2020 – Aug 2023
  • Led modernization of parliamentary web platforms and designed a testing protocol for high-stakes congressional sessions (VP, ministers, MPs), raising pre-meeting error discovery to ~95%.
Systems Librarian Intern
Prosentient Systems Pty — Sydney, Australia
May 2019 – Jul 2019
  • Developed a Java-based real-time communication tool improving cross-institutional research collaboration.
  • Administered Linux servers and open-source platforms (DSpace, Openfire) using Git and SSH.
Software Developer, Technical Support Specialist
National Assembly of Zambia — Lusaka, Zambia
Jun 2012 – Aug 2018
  • Provided technical support for 500+ staff and built an automated provisioning script for a country-wide network, cutting configuration time 90% (100% consistent) and issue resolution time 30%.

Education

Ph.D. in Computer Science — Software Engineering & Cybersecurity
Idaho State University — Pocatello, ID  ·  GPA: 4.0
Aug 2023 – May 2027
  • Dissertation: Interactive IDE-integrated software visualization tools for code comprehension
  • Research areas: Empirical Software Engineering, AI-assisted development, Software Supply Chain Security
M.Sc. in Computer Science
University of New Orleans (now LSU New Orleans) — New Orleans, LA
2018 – 2020
B.Sc. in Computer Science
University of Zambia — Lusaka, Zambia
2007 – 2011

Selected Publications

6 peer-reviewed publications at top software-engineering venues. Also presented at ITNG 2025 (Springer) and i-ETC 2024/2026; full list with citation counts on Google Scholar →

ICSME 2026 · IEEE International Conference on Software Maintenance and Evolution
Comprendia: AI-Augmented Code Comprehension
Nachuma, C. & Zibran, M. F.
Tool Demonstration and Data Showcase Track
arXiv:2608.10290
MSR 2026 · 23rd International Conference on Mining Software Repositories
When AI Teammates Meet Code Review: Collaboration Signals Shaping the Integration of Agent-Authored Pull Requests
Nachuma, C. & Zibran, M. F.
doi.org/10.1145/3793302.3793561
MSR 2025 · 22nd International Conference on Mining Software Repositories
Decoding Dependency Risks: A Quantitative Study of Vulnerabilities in the Maven Ecosystem
Nachuma, C., Hossan, M. M., Turzo, A. K. & Zibran, M. F.
doi.org/10.1109/MSR66628.2025.00048
MSR 2024 · 21st ACM International Conference on Mining Software Repositories
ChatGPT in Action: Analyzing Its Use in Software Development
Champa, A. I., Rabbi, M. F., Nachuma, C. & Zibran, M. F.
doi.org/10.1145/3643991.3645077
ACM SAC 2024 · 39th ACM/SIGAPP Symposium on Applied Computing
SBOM Generation Tools Under Microscope: A Focus on the npm Ecosystem
Rabbi, M. F., Champa, A. I., Nachuma, C. & Zibran, M. F.
doi.org/10.1145/3605098.3635927
ACM CompEd 2025 · ACM Global Computing Education Conference
Future Ready, Globally Relevant: Aligning Africa's Tertiary CS Education with Global Standards
Jonathan, M., Oudshoorn, M., Binamungu, L. P., Bainomugisha, E., Nachuma, C., et al.
doi.org/10.1145/3736251.3749527

Technical Skills

Languages
PythonJava C#JavaScriptSQL
Systems & DevOps
GitDocker GitHub Actions (CI/CD)Linux JGitEclipse JDT
Data & Databases
PostgreSQLMySQL OracleNeo4j (Cypher/graph queries) DuckDBParquet
LLM & AI Integration
Anthropic ClaudeOpenAI OllamaPrompt/Context Engineering Graph-Grounded Retrieval
Automation & Data Pipelines
REST API IntegrationBeautifulSoup ETL ScriptingPandasNumPy
Research & Statistical Methods
Logistic Regression (clustered SEs)Wilson Score CIs Systematic MappingQualitative Coding
Cybersecurity
Supply Chain RiskThreat Intelligence SBOM EvaluationCVE/OSV Analysis

Awards & Recognition

🏆
Best Oral Presentation
ISU Research & Creative Works Symposium, 2026
🥇
Best Poster Award
ISU Research & Creative Works Symposium, 2026
🥇
Best Poster Award
4th Annual Intermountain ITC Conference, 2024
🏆
Best Oral Presentation
ISU Research & Creative Works Symposium, 2024

Leadership & Volunteering

Team Lead — TechEquity AI Summit, Plug and Play Tech Center, Sunnyvale, CA  ·  Nov 2025
Led 50+ volunteers supporting 2,000+ attendees; coordinated exhibitor and executive stakeholder engagement.
CS Mentor — Idaho State University  ·  2023 – Present
Provide mentorship in programming fundamentals, career development, and technical interview preparation.
President — University of Zambia Computer Society (UNZACS)  ·  2010
Organized coding bootcamps and workshops, increasing student participation in computing initiatives.
Certifications
🎓 IBM / Coursera: Python for Data Science, AI & Development (2026)

Let's Connect

Whether you're a collaborator, fellow researcher, or just want to connect — feel free to reach out.