MiDAS (Mission Director and Autonomy Systems) Software Engineer Intern
- Engineered AI computer vision solutions in C++, creating a custom neural network with backpropagation (without external libraries), improving object detection accuracy to 99.98% across 10,000+ test images.
- Implemented OTSU thresholding methods to enable object detection without prior model training with a 95% success rate, ensuring backwards compatibility with the neural network.
- Created an internal universal debugger using binary and function instrumentation, used to speed up fault analysis.
- Applied SYS-ML modelling in line with the UK MoD Pyramid Reference Architecture and contributed in Agile sprints using Jira (planning, stand-ups, retrospectives).
- Presented technical deliverables to senior leadership, including the CTIO and liaised with the CFO, gaining exposure to C-level decision-making and strategic project discussions.
C++
AI/ML
Computer Vision
SYS-ML
Agile
Jira
MiDAS