Börkur Sigurbjörnsson is an information scientist and data engineer with expertise in structuring knowledge and extracting insights from data, predominantly in the context of early stage startups.
He has experience working on applied research and prototyping within academia and large corporate research organizations, as well as being a key member and essential contributor of early-stage startups in Iceland, Netherlands, United Kingdom, Spain, Germany and France.
With thorough understanding of academic research and customer development, Börkur transfers state of the art science into innovative products.
Service offering includes
Fine-tuning of task specific large language models.
Training and evaluation of machine learning algorithms.
Training and evaluation of search engine ranking algorithms.
Past work (selected examples)
Fine-tuning of large language models (LLMs) to detect and extract information about the impact of urban interventions (Urbanixm).
Machine learned estimation of reading level needed to understand a piece of information (WizeNoze).
Machine learned identification and disambiguation of brand references in media articles (brandcrumb)
Börkur has a strong track record building prototypes and MVPs within research organizations and startups, for rapid validation of business hypotheses.
Service offering includes
Rapid processing of open-data to create quick proofs of concepts.
Simple web-based information exploration interfaces.
Coordinating engineering work and prioritization for agile delivery.
Past work (selected examples)
Exploration of evidence about the impact of urban interventions and policy (Urbanixm)
Exploration of Flickr photos using semantically structured tag clouds [summary | technical report] (Yahoo! Research)
Focused access to information from Wikipedia by a search engine going below the article level [DIR 2006 | pdf] (University of Amsterdam)
Börkur combines his skills in data science and fiction writing to generate engaging narratives around data.
Service offering includes
Extracting insights from structured and unstructured data sources.
Visualization of data for understanding patterns and opportunities.
Narrated summaries of large data sets.
Past work (selected examples)
Exploration of the impact of urban interventions (Urban Data Stories).
Market pricing of energy performance certificates (freelance).
Finding opportunities for symbiotic CO2 recycling (freelance).
Börkur has implemented diverse pipelines for making heterogeneous data interoperability through centralized data-stores and APIs.
Service offering includes
Focused web-crawls and data extraction/scraping.
Transform heterogeneous data into a centralized data format.
Search APIs for large text collections.
Past work (selected examples)
Focused crawling and scraping of websites related to urbanism (Urbanixm).
Discovery of diverse data sources related to commercial real estate and ingestion into an interoperable data-format (Placemake.io).
Knowledge graph generation from diverse sources for use in faceted exploration of image search results [WWW 2010 | pdf] (Yahoo! Research / Yahoo! Image Search).
Get in touch on LinkedIn or drop me an email for more information about prices and availability.