• Designed, developed and custom trained deep learning models (CV and NLP) that power a Digital Trade Operations solution used to detect, recognise, and extract key information from documents. It is deployed on Microsoft Azure and is currently used in two continents.• Continuously optimise and finetune AI models and extractions scripts to improve performance, accuracy, and scalability.• Part of a team which includes data scientists, cloud engineers, and full stack developers, building… Show more • Designed, developed and custom trained deep learning models (CV and NLP) that power a Digital Trade Operations solution used to detect, recognise, and extract key information from documents. It is deployed on Microsoft Azure and is currently used in two continents.• Continuously optimise and finetune AI models and extractions scripts to improve performance, accuracy, and scalability.• Part of a team which includes data scientists, cloud engineers, and full stack developers, building scalable AI solutions across Ab InBev Nigeria and Ab InBev Mexico.• Optimised data annotation solution for cleaning, labeling, and preparing unstructured data used to train OCR models. Show less

• Performed extensive data exploration, wrangling, and visualization to ensure data quality and relevance for ML models.• Designed and deployed machine learning model for predicting liver disease using Python and Flask.• Participated in a team that competed in the Longhack hackathon and ranked as one of the top 13 teams.• In-depth data analysis on longevity companies to reveal insights, and patterns which include industry's momentum and funding.

• Designed, developed and trained machine learning model to analyze data from 400,000+ customers to predict churn rate within 90 days. • Performed extensive data curation, data cleaning for training and vali- dation for ML models• Implemented an XGBoost model to analyze loan defaulters with 84% accuracy.• Developed machine learning models for predicting daily bike rentals using Python and Flask.• Conducted sentiment analysis on 40,000+ tweets using a Long Short Term Memory… Show more • Designed, developed and trained machine learning model to analyze data from 400,000+ customers to predict churn rate within 90 days. • Performed extensive data curation, data cleaning for training and vali- dation for ML models• Implemented an XGBoost model to analyze loan defaulters with 84% accuracy.• Developed machine learning models for predicting daily bike rentals using Python and Flask.• Conducted sentiment analysis on 40,000+ tweets using a Long Short Term Memory (LSTM) NLP model• Utilized version control systems (GIT). Show less

• Utilised analytical expertise to provide insights and proposals to support business improvements• Extracted and analysed 500+ lines of data on fuel consumption, to help the company reduce cost and maximise profit • Reviewed 300+ records for deficiencies, corrected any incompatibilities and checked outputs using Microsoft Excel• Maintained records, receipts, and withdrawals of office items which helped to reduce wastage in the office