About

I build language technology for under-resourced African languages and data systems that hold up in regulated, high-stakes settings like health and pharmaceutical research.

I am a volunteer researcher with Masakhane, the open research community for African NLP. There I have fine-tuned multilingual transformer models in PyTorch and Hugging Face, prepared pretraining and evaluation data, and annotated Yoruba data for cross-lingual retrieval. I co-authored AfroLM (SustaiNLP @ EMNLP 2022), CIRAL (ACM SIGIR 2024), and AfriMTE & AfriCOMET (NAACL 2024).

Alongside research, I have spent more than eight years in data roles. I currently work as a Documentation Automation Engineer on assignment at Regeneron Pharmaceuticals, where I run production data workflows with CI/CD, test harnesses, and data lineage controls, and use NLP/LLM tooling to generate traceable scientific reports. Before that I worked on HIV surveillance data reconciliation at Health Research, Inc. in New York, and as a data analyst at Nigeria's Defence Space Administration.

I hold an M.S. in Computer Science from The College of Saint Rose and a B.S. in Computer Science from Joseph Ayo Babalola University, Nigeria. I also founded the EquiHealth Institute, an independent research initiative on healthcare access.

News

  • Starting my PhD at Simon Fraser University with Prof. Marzena.
  • Filed a U.S. provisional patent application on disease outbreak forecasting.
  • Co-patentee on a Nigerian patent for an AI-driven drug repurposing system. Founded the EquiHealth Institute.
  • Joined Insight Global as a Documentation Automation Engineer, assigned to Regeneron Pharmaceuticals.
  • CIRAL, a cross-lingual retrieval test collection for four African languages, presented at ACM SIGIR 2024.
  • AfriMTE & AfriCOMET presented at NAACL 2024.
  • Completed my M.S. in Computer Science at The College of Saint Rose.
  • AfroLM presented at the SustaiNLP workshop at EMNLP 2022.

Research Interests

  • NLP for African languages

    Pretraining, fine-tuning, and evaluating multilingual models for languages such as Yoruba, Hausa, Somali, and Swahili, where labelled data is scarce.

  • Data-efficient language models

    Active learning and other methods that get strong results from far less training data than standard baselines.

  • Evaluation & retrieval

    Machine translation metrics and cross-lingual information retrieval test collections built with native-speaker judgments.

  • AI for health

    Forecasting, drug repurposing, and trustworthy data pipelines for public health and biomedical research.

Publications

For the complete and current list, see Google Scholar.

  1. SIGIR2024

    CIRAL: A Test Collection for CLIR Evaluations in African Languages

    …, O. O. Awoyomi, …

    Proceedings of the 47th International ACM SIGIR Conference on Research and Development in Information Retrieval

    A cross-lingual retrieval test collection with 1,600+ English queries, 2.5M+ passages in Hausa, Somali, Swahili, and Yoruba, and about 30k relevance judgments by native speakers. I annotated the Yoruba data.

    Information retrievalYoruba
    Find paper
  2. NAACL2024

    AfriMTE and AfriCOMET: Empowering COMET to Embrace Under-resourced African Languages

    J. Wang, D. I. Adelani, S. Agrawal, …, O. O. Awoyomi, …

    Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics (NAACL-HLT), pp. 5997–6023

    A human-annotated machine translation evaluation dataset for African languages (AfriMTE) and COMET-based metrics adapted to them (AfriCOMET).

    Machine translationEvaluation
    Find paperarXiv
  3. SustaiNLPEMNLP 2022

    AfroLM: A Self-Active Learning-based Multilingual Pretrained Language Model for 23 African Languages

    B. F. P. Dossou, A. L. Tonja, O. Yousuf, S. Osei, A. Oppong, I. Shode, O. O. Awoyomi, C. C. Emezue

    Proceedings of the Third Workshop on Simple and Efficient Natural Language Processing (SustaiNLP), EMNLP 2022

    A multilingual language model pretrained from scratch on 23 African languages with self-active learning. Trained on about 14× less data than existing baselines, it outperforms AfriBERTa, XLM-R-base, and mBERT on NER, text classification, and sentiment analysis. I prepared training data, annotated evaluation data, and ran experiments.

    PretrainingActive learning
    Find paper

I have also co-authored work at IJCNLP-AACL 2023 and journal articles on health data and AI. Full titles and author lists are on my Google Scholar profile.

Patents & Projects

  • AI-driven drug repurposing system

    Nigerian business method patent, co-patentee, 2025.

  • Disease outbreak forecasting

    U.S. provisional patent application, inventor, filed 2026 (not yet examined).

  • LinguaCast: public health forecasting prototype

    Forecasting models on 711 weeks of CDC FluView ILINet influenza-like illness data (2010–2024). The linear regression model reached an out-of-sample R² of 0.9927 on held-out weeks. Research prototype only.

  • EquiHealth Institute

    Founder of an independent research initiative on healthcare access, 2025–present.

Experience

  1. 2025 – present

    Documentation Automation Engineer

    Insight Global, assigned to Regeneron Pharmaceuticals

    Production data workflows with CI/CD and test harnesses; ETL for microscopy, flow cytometry, and sequencing data; NLP/LLM-supported report generation with lineage and provenance.

  2. 2022 – present

    NLP Researcher (Volunteer)

    Masakhane open research community

    Fine-tuning multilingual transformers on GPU clusters; co-author of AfroLM, CIRAL, and AfriCOMET.

  3. 2024

    Data Administrator

    Health Research, Inc., Albany, NY

    Reconciled HIV surveillance records between the CDC's eHARS and New York's NYEHMS; documented data import code. SQL, SAS.

  4. 2024

    Lead Programmer / Analyst

    The State University of New York

    PL/SQL forecasting models and Tableau dashboards for university planning.

  5. 2022 – 2023

    Graduate Assistant

    The College of Saint Rose

    Secure document storage and retrieval with SQL FileStream and SMB.

  6. 2018 – 2021

    Space Science & Technology Officer, Data Analyst

    Defence Space Administration, Abuja, Nigeria

    Automated reporting with Python and ArcGIS.

  7. 2016 – 2018

    Data Analyst

    GBIT Consulting Limited, Abuja, Nigeria

Education

  1. 2027 –

    Ph.D.

    Simon Fraser University, Burnaby, BC · Supervisor: Prof. Marzena

  2. 2022 – 2023

    M.S. in Computer Science

    The College of Saint Rose, Albany, NY

  3. 2010 – 2014

    B.S. in Computer Science

    Joseph Ayo Babalola University, Nigeria

Skills

Machine learning

PyTorch, Hugging Face Transformers, vLLM, multilingual fine-tuning, GPU training

Data engineering

ETL, automated validation, CI/CD, Git, data lineage, schema governance

Programming

Python, SQL, PL/SQL, SAS, Flask

Databases & BI

PostgreSQL, SQL Server, MySQL, Tableau, Power BI, ArcGIS