Chiamaka Mary Igwilo
Ames
Summary
Machine Learning Engineer & PhD Researcher specializing in LLM agent evaluation, trace-based debugging, and failure-mode analysis. Experienced in building and deploying AI/ML systems using Python, PyTorch, NLP, and large-scale data pipelines, with hands-on work in HPC and distributed systems and practical experience with models and frameworks including Llama, LangChain and Hugging Face.
Overview
9
9
years of professional experience
Work History
RESEARCH ASSISTANT – LLM Agent Failure Analysis
Iowa State University
Ames
08.2023 - Current
- Built a Python-based evaluation toolchain using LangChain, CrewAI, OpenAI API (GPT) and HF models to collect and analyze multi-agent execution traces.
- Identified 12+ recurring failure modes across GPT, Llama, and Mistral models by comparing expected vs. observed agent behavior.
- Produced reliability benchmarks and diagnostic reports summarizing failure frequency and root causes across agent frameworks.
- Refined prompts and workflow logic based on trace insights, improving task consistency across repeated runs.
TEACHING ASSISTANT - Python & Data Science
Iowa State University
Ames
08.2023 - Current
- Delivered Python, data science, and applied ML instruction to 300+ undergraduate and graduate students, emphasizing practical implementation and evaluation.
- Designed and implemented hands-on labs on data validation, evaluation metrics, and reproducible ML workflows.
- Reviewed and debugged student code at scale, reinforcing best practices in testing, documentation, and reproducibility.
COMPUTING SCHOLAR
Lawrence Livermore National Laboratory
Livermore
05.2024 - 08.2024
- Built and optimized linear solvers for large-scale scientific simulations on HPC systems using PETSc and HYPRE.
- Resolved numerical instabilities via residual and convergence analysis, improving solver robustness.
- Benchmarked LU, Jacobi, and GMRES methods to accelerate distributed matrix computations.
- Tested solver accuracy and stability and integrated them into real HPC simulation workflows.
DATA SCIENTIST
Natview Technology
Jabi
01.2023 - 07.2023
- Built pipelines and predictive models across five projects, reducing processing time and accelerating analysis.
- Applied statistical modeling, feature selection, and forecasting to improve model accuracy and reliability.
- Worked with financial and health datasets from the statistics bureau, focusing on data quality and reproducibility.
- Maintained integrated data models supporting analytics for three organizations.
MACHINE LEARNING AND DATA ENGINEER
Techxagon
Wuse 2
02.2021 - 12.2022
- Trained and deployed ML models for NLP and predictive analytics using Python and scikit-learn.
- Set up data pipelines and evaluation metrics for multiple projects.
- Cleaned and prepared data for modeling and analysis.
- Worked on NLP tasks including extraction, QA, and sentiment analysis.
IT SUPPORT ENGINEER
Digital Bridge Institute
Abuja
01.2017 - 11.2019
- Installed and supported 200+ Linux and Windows workstations for training environments.
- Provided IT support and troubleshooting for OS, application, and hardware issues.
- Managed endpoint security using McAfee antivirus for multiple organizations.
- Supported basic network infrastructure and trained users on common software tools.
Education
DOCTOR OF PHILOSOPHY - COMPUTER SCIENCE
Iowa State University
05-2028
MASTER OF SCIENCE - COMPUTER SCIENCE
African University of Science and Technology
02.2022
BACHELOR OF SCIENCE - COMPUTER SCIENCE
Godfrey Okoye University
07.2016
Skills
- OpenAI API
- Hugging Face Transformers
- LangChain
- CrewAI
- Pinecone
- Llama
- RAG
- NLP
- LLMs
- Model Evaluation
- Error Analysis
- Prompt Engineering
- Python
- PyTorch
- NumPy
- Pandas
- SQL
- Git
- C
- AWS
- Parallel Computing
- PETSc
- HYPRE
- Numerical Solvers
- V&V
- Verification and Testing
- Traceability
- Reliability Analysis
- Ethics in AI
Publications
- Comparative analysis of ensemble learning and non-ensemble machine learning algorithms for phishing URL detection, Igwilo, C. M., & Odumuyiwa, V., FUOYE Journal of Engineering and Technology, 7, 3, 305-312, 2022, https://doi.org/10.46792/fuoyejet.v7i3.807
- Automatically detecting numerical instability in machine learning applications via soft assertions, Sharmin, S., Zahid, A. H., Bhattacharjee, S., Igwilo, C. M., Kim, M., & Le, W., Proceedings of the ACM on Software Engineering (FSE 2025), 2806-2827, 2025, https://doi.org/10.1145/3729394
Timeline
COMPUTING SCHOLAR
Lawrence Livermore National Laboratory
05.2024 - 08.2024
RESEARCH ASSISTANT – LLM Agent Failure Analysis
Iowa State University
08.2023 - Current
TEACHING ASSISTANT - Python & Data Science
Iowa State University
08.2023 - Current
DATA SCIENTIST
Natview Technology
01.2023 - 07.2023
MACHINE LEARNING AND DATA ENGINEER
Techxagon
02.2021 - 12.2022
IT SUPPORT ENGINEER
Digital Bridge Institute
01.2017 - 11.2019
MASTER OF SCIENCE - COMPUTER SCIENCE
African University of Science and Technology
BACHELOR OF SCIENCE - COMPUTER SCIENCE
Godfrey Okoye University
DOCTOR OF PHILOSOPHY - COMPUTER SCIENCE
Iowa State University