Eunico Cornelius

• Developed in an agile environment team an Intelligent Attendance System that records attendance using image recognition and detects covid symptoms using thermal detection with Python (Flask), ZMQ, Redis, MySQL, Quasar and VueJS. • Developed E2E CI/CD workflow using bitbucket pipelines with Cypress for automation testing per container. • Product bought by ShapesAI and have been used for COVID awareness in London, broadcasted on BBC.

• Built machine learning model that converts images of electricity meters (PLN) into digital form with 80% accuracy using 1 million+ self annotated image data. • Built backend server REST API to serve the model above and company projects using MongoDB, Werkzeug, Bcrypt, and FlaskJWT. • Conducted user load testing on docker swarm using locust and multiple docker containers with 6 nodes (1 manager, 5 workers).

• Developed end-to-end machine-learning app using Python, Tensorflow, and Keras, converting physical photo IDs into digital text. • Singlehandedly built a self-developed CRNN neural network with an accuracy of 95% that is the central piece of the photo ID app. • Created the web interface for the machine learning model in Flask / HTML enabling users to use the ML tool in real-time. • Submitted a white paper with a team of three, containing technical findings of the neural network project.

• Built and optimized the front-end features for the company website using HTML/CSS, PHP, JS with Processwire -- https://Yamdu.com. • Developed live newsfeed feature for the company, displaying recent news and other web-app information -- https://Deinkinoticket.de. • Created and trained a chatbot in Facebook Messenger that allows customers to purchase cinema tickets via Facebook.

Development of WEB MVC using Visual Basic (.NET) and SQL Server 2008

Machine Learning, Software Development