Aidan Kubeyev

• 5 years of machine learning experience in a startup and academia in the UK, focusing on a product
• 3 years in research software development and 2.5 years in predictive modelling in global consulting

Scientist in Computation AI/ML –  Concr.coNov 2021 – July 2024Cambridge, United Kingdom

• Developed and deployed a machine learning method in production that helped Concr secure its first commercial client, marking a significant startup milestone
• The Clinical Trials Machine Learning Predictor - as part of a team, I developed a pioneering machine-learning method that simulates clinical trials and recommends targeted treatments for patients using diverse data: genetics, histology, clinical, and image data. Manuscript submitted to Cancer Discovery
• Computer Vision for Medical Pathology project - I developed a deep learning model to predict human cell types, (e.g. cancer cells or lymphocytes) based on medical pathology images and investigated the impact on drug response. Presented at the AACR conference
• I was the first team member to present orally the machine learning work I led at the AACR conference

• Hybrid Deep Learning and Physics-based Method: I designed and developed a deep learning computer vision project that uses convolutional neural networks and physics-based modelling to create a novel hybrid model. Published in the Engineering Applications of Artificial Intelligence journal 
• Computer Vision Physics-based Research Software: I developed research software and a method that inputs images, performs predictive modelling, and identifies key performance indicators (KPIs) for Energy and Climate Change projects. Published in Transport in Porous Media
• Applied automatic machine learning for predicting the semi-analytical equation when prediction fails 
• PhD thesis: “Development of computational modelling methods for solid and fluid mechanics in fractures …” 

• Completed 4 data-driven predictive modelling projects, applying advanced statistical techniques and physics-based models to optimize industrial processes and improve operational efficiency
• These consulting projects include: a North Sea heavy-oil field simulation study, oilfield production optimization, artificial lift feasibility, and water management numerical study. 
• Utilized Bayesian statistics to analyze large multi-modal datasets and develop predictive models for oilfield performance
• Delivered 30+ client-facing presentations, developing communication and storytelling skills and the ability to translate complex ideas into actionable insights for clients

• Managed and analyzed large-scale production datasets to enhance oilfield performance and efficiency
• Developed and implemented predictive models to forecast production rates, identify anomalies, and detect trends, significantly improving data-driven decision-making processes

Heriot-Watt University – PhD in Computational Science / Machine Learning

• Awarded 210,000 USD James Watt Scholarship and EU Horizon 2020

• Coursework - Merit (68%), Research project - Distinction (77%)

• Distinction (98%)