About Me
I build things
that think.
I build agentic AI systems at NatureAlpha, designing and deploying LLM-powered pipelines for sustainable finance data. Previously, I was a Research Engineer at dunnhumby working on customer data science, and a Research Assistant at Singh Lab using deep learning to study ageing biology. I hold an MSc in Data Science and a BA in History from Brown University.
MSc Data Science
Brown University · 4.0/4.0
BA History (Data Science minor)
Brown University · 3.6/4.0
Education
Brown University · Providence, RI
MSc Data Science
- Singh Lab Research Assistant
- Sustainable Food Initiative
- Brown Club Lacrosse: Won D2 regionals, placed 5th nationals
BA History (Data Science minor)
- Data Science Club Team Lead
- Data Science Fellow
- Research Assistant, ML for Earth & Environment
Published Research
TimeFlies
TimeFlies: an RNA-seq ageing clock for Drosophila
Tennant, N.*, Okonkwo, A.*, O'Connor-Giles, K., Larshen, E., & Singh, P.
Developed a deep learning ageing clock achieving 95% accuracy and 0.99 AUC on Drosophila age prediction. Identified ageing biomarker genes with in vivo validation.
Experience
LLM Engineer
Building agentic AI systems for sustainable finance. Designing and deploying LLM-powered data pipelines, RAG systems, and multi-agent workflows for ESG and biodiversity data analysis.
Research Engineer
Developed NLP and LLM systems for data insight synthesis. Contributed to core functionality involving vector embeddings, cost modelling, impact prediction, and dynamic product ranking.
Associate Research Engineer
Built a complete ML pipeline using autoencoders on a 5M customer dataset for Tesco Mobile handset assortment. Delivered projects generating £829K in business value. Refactored codebases and removed external API dependencies.
Research Assistant
Conducted X chromosome dosage compensation research as part of the NSF-funded IISAGE initiative. Developed computational models for cross-species ageing biomarker analysis across 11 research laboratories.
Projects
Premier League Projections
ML pipeline analysing 457 data points across Premier League seasons using LightGBM, Random Forest, and SVC with SHAP interpretability.
PythonLightGBMRandom ForestSHAPSenID - Senescence Detection
Computer vision for identifying senescent cells using nuclear morphology. Achieved 0.96 AUC with 1/100th the training samples. 2nd place, hackathon (225+ participants).
PythonComputer VisionDeep LearningCNN
Skills
Get in touch
Open to opportunities in AI engineering, research collaborations, and interesting conversations.