Börkur Sigurbjörnsson is an applied scientist and software engineer with a proven track record in bootstrapping product development in early-stage startups—turning ideas into action by rapid prototyping.
Börkur works along the full product stack from back-end to front-end, with a strong focus on back-end services, processing and structuring data and exposing it through APIs.
Data and information harvesting from the Web. Ingestion pipelines into wide range of storage systems. Structuring of data using machine learning.
Designing data models for structured and semi-structured data. Deploying SQL and noSQL data storage instances, search engines and serving data via APIs.
Robust delivery of software by breaking complex systems into clearly specified components. Agile management of development teams through clear alignment of priorities and work deliverables.
Full-stack development of a knowledge base for urbanism. Machine learned classification of urbanism content and a custom search engine for urban professionals.
Design of a generic data model for urban data, harvesting and ingestion of a selection of real-estate related data, and its exposure through REST APIs.
Development and evaluation of machine learned methods for estimating the reading level needed to understand a piece of text. Search engine ranking and associated APIs for education texts for children, taking estimated reading level into account.
Segmenting photo annotation tags into semantic groups using classifiers trained on Wikipedia and WordNet. Prototype for browsing photos through semantic tag clouds.