Bernhard Barker

Real-time automated shift planning for chauffeur service using genetic algorithms, linear regression, random forest and clustering in Python, Go and SQL in Linux with Kafka and AWS. - Discuss and implement ML models for profit, cancellations and traffic - Created text classifier for customer requests with 95%+ precision - Analysed data to find most promising cities to expand to - Developed system end-to-end to generate shifts with up to 53% profit - Investigated production issues and implemented solutions

Mobile role-playing game development using C# with Unity in Windows.

Image classification to filter inappropriate images from Image Search using C++ and Python with MapReduce in Linux. - Developed image classifier with 90%+ precision using neural networks - Developed text classifier to categorise search queries - Launched machine learning models in production - Fixed and extended tool to remove images from Google’s index

Delivery route generation for complex traveling salesman variant using C++ and SQL Server in Windows. - Rewrote clustering algorithm to generate initial delivery routes - Extended hill climbing algorithm which optimises generated routes - Full-stack development of application used by market leaders

Maintenance of transaction processing system of a leading national bank using C, C++, MySQL and PHP in Linux and VOS. - Developed and maintained core applications of financial system - Investigated and resolved production issues in 24/7 support role

Final Year Distinctions: Artificial Intelligence, Computer Networks, Compiler Construction, Discrete Structures, Geometry. Other Final Year Subjects: Computer Graphics, Database Systems, Software Engineering, Trends in Information Technology (Security)