LinkedInLabelled trace dataset
Coding Agent Failure Atlas
A synthetic atlas of coding-agent failure traces for monitor research, with evidence spans, intervention points, and safer counterfactuals.
View GitHub repository
Research
Auditable AI research with cultural context.
Coding-agent evaluation, multilingual NLP, music emotion analysis, low-resource language technologies, responsible AI, and digital heritage research.
Research profile
Agent behaviour, emotion, language, and meaning.
Oyinkansola Onwuchekwa is an AI Engineer and Research Scientist whose work connects coding-agent evaluation, low-resource NLP, responsible AI, and cross-cultural emotion modelling. She builds tested Python tools, trace datasets, monitor harnesses, and evidence-grounded workflows for research and decision-making.
Her PhD research at the University of Hull explores cross-cultural musical elements, emotional expression, and genre characteristics in contemporary global music lyrics using NLP and deep learning. The wider research profile sits at the intersection of AI, music, language, identity, cultural heritage, and public good.
Public research work
Research code, released tools, and maintainer-reviewed contributions.
Public work across coding-agent evaluation, low-resource NLP, and open-source AI tooling.
View GitHub repository
Monitor evaluation harness
Coding Agent Monitor Lab
A small research lab for testing whether monitors can catch risky coding-agent traces using structured evidence rather than broad completion scores.
View GitHub repository
Released Python package
Low-Resource NLP Toolkit
Utilities for African language pre-processing, emotion-label mapping, evaluation, language routing, and code-switching audits.
Maintainer-accepted contribution
Haystack
Accepted contribution to the Haystack framework, a widely used open-source library for building AI applications.
Maintainer-accepted contribution
Sentence Transformers
Accepted contribution to Sentence Transformers, supporting practical work with embedding models and semantic search.
Research interests
Technical depth shaped by cultural knowledge.
Coding-agent evaluation
Trace-level datasets, monitor experiments, evidence spans, intervention points, safe counterfactuals, and testable failure modes for agentic AI systems.
Multilingual NLP
Low-resource language modelling, language routing, bias-aware evaluation, African language NLP, and multilingual data pipelines.
Emotion and music AI
Emotion recognition in music and text, cross-cultural analysis, lyric modelling, Afrobeats analytics, and multimodal interpretation.
Responsible AI
Human-centred AI, fairness, representation, digital heritage, community-led data governance, and AI literacy in cultural contexts.
RAG and research agents
Systematic literature review workflows, protocol generation, screening support, evidence synthesis, auditability, and reproducible structured outputs.
Roles and recognition
Research, teaching, peer review, and public service.
Public materials
Institutional project context and research writing.
Featured project
DAIL-ICH: Digital/AI literacy for community-led cultural heritage data governance
British Academy-funded research on responsible technology, digital heritage, community knowledge, and AI literacy for intangible cultural heritage contexts.
Medium
Medium article
AI-Powered Cross-Cultural Study of Emotions in Music Lyrics
A research-facing Medium article on AI, music lyrics, emotion analysis, and cross-cultural interpretation.
Read on MediumResearch collaboration
Build culturally grounded AI with KKC.
For academic invitations, applied AI partnerships, reviews, workshops, and research collaborations.
Start a research enquiry