Imperial College London

Ecole Européenne de Bruxelles I

Imperial College PhD physicist, deep learning scientist, entrepreneur

* Lead developer and system designer on event-driven, deep learning-based “intelligent” checkout system for Europe’s 3rd largest supermarket brand, based on Nvidia Jetson edge computing platform.

* Researched and tested various object detection, OCR, and “shallow” models for use in anti-shrinkage ensemble solution.

* Productionised SotA neural networks for automatic licence plate recognition (ALPR) systems with sub-1/10000 error rates.

* Ushered in several improvements in development processes, CICD, coding best practices, and collaboration. Planned, facilitated and oversaw entire team’s transition from subversion to git, started code review culture.

* Acted as bridge between external clients, hardware, software & research teams by translating requirements and issues between stakeholders of varying backgrounds.

Prime Vision

Computer Vision Engineer

* Tech lead and co-founder at Project Lucas (later rebranded to Yepic AI), a Founders Factory incubator start-up specialising in image manipulation and generation for the advertising industry, using generative adversarial networks (GANs).

* Managed and led a team of 3 ML scientists while handling both product and tech roadmaps.

* Designed, built, and (in)validated several ensemble computer vision systems featuring pose estimation, object detection, and geometric transformation, for the purpose of “painting” clothes onto generated people.

* Trained and benchmarked several convolutional neural net (CNN) architectures for semantic segmentation tasks.

* Researched and prototyped several GAN architectures to develop a controllable image generator for fashion imagery.

Project Lucas

Co-Founder & CTO

* Built Nate (www.nate.tech), an AI-driven web automation company. Latest raise $38M in June 2021.

* Researched, built, trained, tested, and tuned LSTM/GRU-based systems for machine translation tasks on computer code.

* Prototyped simple speech-to-text and NLP systems for part-of-speech recognition and sentiment analysis based on recorded voice data from financial sector phone calls.

* Advised both in-house companies and internal investment committees on any and all ML-related matters.

* Performed in-depth product discovery-related research and interviews in order to find product-market fit for incubation project start-ups, used learnings to influence product design as well as strategy and project management decisions.

* Created and developed a relationship between the Factory and Imperial College London’s Entrepreneur Society and career offices, in order to foster cooperation and drive innovation..

* Engaged in making positive changes to both team and company culture, both in terms of technical discipline and general operation -- e.g. started a journal club across accelerator and incubator companies.

* Actively participated in recruitment and screening of CEO and CTO candidates for Factory ventures, as well as technical personnel at all levels.

Founders Factory Ltd

Machine Learning Scientist

* 6th original member (3rd tech) of high-profile, London-based AI startup from early seed stage.

* Performed topic, emotion and sentiment analysis on streaming, user-generated Twitter data using novel natural language processing techniques.

* Designed and developed knowledge graph-based ROLAP framework to dynamically combine and query disparate data sources, allowing machine learning to be performed on variables across datasets.

* Architected, built and maintained ETL pipelines for batch-delivered and streaming datasets of terabyte magnitude.

* Designed and built automated geospatial data engineering pipelines using PostGiS, GeoSpark, and Airflow.

* Designed and built Elasticsearch-backed predictive query engine for geographic points of interest in the UK.

* Designed, built and maintained RESTful API endpoints for consumption by web app’s front end, allowing on-demand querying, combination and aggregation of various datasets.

* Implemented federated authentication interface for corporate users of customer-facing SaaS product.

GYANA

* Designed and developed analytics tools to leverage large sets of user-generated data from popular social media platforms, e.g. YouTube, Facebook, Instagram, using scientific Python stack.

* Collaborated with key stakeholders to identify metrics or data sets to explore, in order to better drive decision making and identify possible revenue avenues.

* Delivered regular earnings reports, as well as ad-hoc analysis focussing on data visualization.

* Optimised current revenue reporting systems, from both execution time and automation perspectives.

Brave Bison (formerly Rightster)

Data Scientist

* Nominated for industry award by client for development on big data-driven web application with PL/SQL back end, allowing real-time anti-money-laundering/know-your-customer screening of current and prospective business partners for UK financial firms.

* Designed and implemented full unit testing suite for project code base, in combination with Docker/Jenkins framework, leading to more robust and easily deployable solutions.

* Optimised, documented and reformatted entire project repository during downtime, freeing up time for QA, allowing new colleagues to more efficiently understand and contribute to project, and ultimately increasing team’s output quality, as well as project auditability.

* Actively participated in daily client conference calls and meetings to manage targets, deadlines and expectations concerning regular feature development for data-driven analysis tools.

* Aided in the training of junior colleagues by teaching elements of coding, analysis methods, and knowledge of the data ecosystem.

Diego Andersen