Tom Reader - CV

Computer Vision and Machine Learning Consultant

Work email: tom@alvervalleysoftware.com

To download my CV in PDF form, please click here

Education

BSc Computer Science, Coventry University

• 2:1 (Honours)
• Specialised in AI / Machine Learning
• Final year project focused on Neural Networks and Genetic Algorithms

Recent Experience

Consulting on a number of projects, mostly long-term, including:

• Developed various computer vision solutions for an on-board drone guidance system
• Worked with other members of the guidance system team to develop, assess, integrate and test a number of solutions
• This resulted in a very successful system being developed
• I am named on a patent for my development of part of this system

Retail hardware supplier (details commercially sensitive) – ongoing, since 2023

• Developed computer vision system for classifying items purchased by customers
• Achieved higher accuracy than existing commercial system, and at much higher speed
• Also performed platform-specific development (e.g. portation to high speed processor), and integration to the rest of their product
• Product now rolling out across stores

MachineEye Technology – Computer Vision Consultant – ongoing, since 2020

• Developed and rolled out a safety-compliance assessment system, detecting people in 'danger areas' in industrial environments, and other safety factors
• Assessed and selected hardware, chose software models, used deep learning techniques to build custom models for specific requirements
• Worked alongside other members of the team to roll out the product to a number of customers

Rumex / Taylor Technologies – since 2020

• Lead software developer on a chemical spot-spraying system to control spray of weed-killers on commercial farms, saving around 77% of chemical costs
• Developed deep-neural-network based system that could identify target weeds (docks) very accurately, ensuring valuable plants (clover and grass) are untouched

Work on various other contracts

• Electron microscope image analysis software, microchip / MEMS fabrication quality control analysis software, cosmetics analysis tool, astronomical image processing

Technologies and Expertise

• 'Traditional' computer vision, using libraries such as OpenCV
• Modern computer vision methods, including:
  • Classifiers such as ResNet
  • Detectors such as MobileNet-SSD, Yolo
    • Recent experience of training and deploying deep learning models on 'edge' hardware with very tight run-time constraints
• Machine learning:
  • Neural Networks, Deep Learning, including PyTorch, TensorFlow
  • Genetic algorithms
  • Pre-processing, e.g. data augmentation techniques
  • Training and run-time inference frameworks
• Hardware platforms such as the Jetson range (NX, Orin) and RockChip, as well as Raspberry Pi, OrangePi, Vantron 
• 30+ years' experience in general programming and software development:  C++ and Python, Linux, database, data communications

Work Environment

• I work mostly from home (south coast of England), although occasional on-site meetings by arrangement are always possible
• My work environment is secure:  I have current Cyber Essentials certification, and take client confidentiality very seriously

Hobbies / Interests / Volunteering

• Active member of local running club, also keen cyclist
• Occasional volunteer at parkrun and other events
• Training member of SERV Wessex (the "Blood Bikers") as a volunteer driver
• Recently passed Advanced Driving Test (achieved RoSPA "Gold" standard), been asked to train as a tutor to help future pupils – in progress
• Current DBS held with various organisations