Looking for a job

Omokemi Ayodimeji

@omokemiayodimeji

Message

I build efficient LLM reasoners and production-ready ML systems.

Nigeria

Message

What I'm looking for

I'm looking to build efficient, reliable ML and LLM systems across data preparation, evaluation, deployment, and human-in-the-loop decision workflows.

I fine-tuned Gemma-3-270M for multi-hop reasoning in Hop-Specialist, publishing the model, dataset, and live Hugging Face demo. I also built a hybrid email-triage system where policy rules determine automatic routing or human escalation.

At OSOPADEC, I improved budget-allocation forecast accuracy by 25% with XGBoost and SARIMA and automated SQL validation to reduce reporting errors by 40%. I also research contactless fingerprint matching at Queryfier and teach applied ML, deep learning, NLP, and Python at CAMLDS.

Experience

Work history, roles, and key accomplishments

Current

Machine Learning Engineer

Current

Queryfier

Sep 2025 - Present (1 year)

Researched contactless fingerprint matching, building a classical minutiae-based baseline with OpenCV and NIST NBIS tools, achieving 19.05% EER. Teaches core ML at CAMLDS to more than 10 students, covering model training, evaluation, applied deep learning, NLP, and Python.

OpenCV NIST NBIS EER TAR@FAR Python Machine Learning Deep Learning NLP Curriculum Development

Data Scientist

OSOPADEC

Mar 2024 - Jan 2025 (10 months)

Built XGBoost and SARIMA models for budget allocation, improving forecast accuracy by 25%. Automated data validation pipeline, reducing reporting errors by 40% and cutting report generation time from 5 days to 3. Led a cross-functional team of 10 to streamline resource-allocation workflows, saving over 200 labor hours quarterly.

XGBoost SARIMA SQL Time series Forecasting Data Validation Team Leadership