Robert Vava

Founding engineer who built frontend and backend infrastructure for asothia.com, maintained a rapidly growing user base, and developed web scraping agents with 92% accuracy and matching algorithms using vector DBs and LLMs.

Built an optimized NoSQL system for multi-session neuronal spike data, implemented a modified transformer and graph-based post-processing to boost neuron tracking accuracy by 18%, and acted as main codebase and data manager for a team of 7.

Developed an embedded traffic analysis and control system with object detection and risk assessment, deployed edge ML models on millimetre-wave and video data, and implemented an Azure IoT data streaming pipeline and dashboards.

Built a machine learning system that increased supporter reach-out donations and volunteering ROI by 13x and delivered data science reports, dashboards, and backend systems under tight deadlines.

Trained and deployed predictive models to improve decision making for financial, scientific, and FMCG clients, and instructed Level 4 industry apprentices in Data Analyst and Software Engineering programmes.

Delivered workshops, labs, and marking for modules in mathematics, computer vision, machine learning, and dynamical systems, and assembled programmed robots and scientific simulations for department activities.

Researched multi-kernel ML models for anomaly detection in multimodal TBI datasets, co-authored a systematic review/meta-analysis, and contributed to software tools for neuroimaging analysis.

Grade: Merit

Activities and societies: Dissertation supervised by Dr. Ivor Simpson; worked with variational/adversarial autoencoders, GANs, and diffusion models.

MSc Artificial Intelligence and Adaptive Systems with a dissertation recreating visual stimuli from brain signals using autoencoders and generative models; awarded Merit and Chancellor's Scholarship for Academic Performance.

Grade: First Class

Activities and societies: Research project measuring EBR and development of Python/Raspberry Pi/C++ detector.

BSc Psychology with Education focused on researching eye blinking rates as a proxy for dopaminergic activity; graduated First Class and developed an automated EBR detector.

Completed Harvard's CS50x Introduction to Computer Science online course to gain foundational programming and computer science skills.

Completed an Introduction to Molecular Spectroscopy certificate course to supplement knowledge in molecular measurement techniques.