Chief Architect
CurrentA tech start-up using machine learning to automatically clean and enrich large volumes of business data quickly and effectively. - Established the back-end architecture, hosted on AWS, based on microservices connected by an API, deployed as containers/lambda functions. - Integrated SSO (OAuth/OIDC) to secure the company’s applications and API. - Developed and deployed machine learning models for data cleaning (Python/TensorFlow): # Models including classification, entity recognition and semantic similarity. # Techniques including transformers, CNNs, word vectors/embeddings and transfer learning. - Published an Excel Add-In (JavaScript/React) to utilize and train the above models directly in Excel, in real time. - Developed a python package (‘Wrangles’) to encapsulate the above capabilities for integration into company and third-party systems.