Thomas Dybdahl Ahle work email
- Valid
- Valid
- Valid
Thomas Dybdahl Ahle personal email
- Valid
Machine Learning Scientist at Normal Computing, where I work on auto-formalizing systems for advanced hardware. Previously, I was a Research Scientist at Meta, focusing on scaling language models and ranking algorithms. I completed my PhD in Locality Sensitive Hashing at the University of Copenhagen (2019), where I also co-founded an award-winning NLP startup as Chief of ML. My research interests span algorithms, machine learning, and software engineering, with a particular emphasis on large-scale AI systems and efficient data structures. I have a broad background, having done previous work on algorithmic optimization, natural language processing, and scalable machine learning architectures.
-
Head Of Ml, Research ScientistNormal Computing May 2023 - PresentNew York, Ny, Us -
Research ScientistMeta Mar 2020 - May 2023Menlo Park, Ca, UsI lead the Machine Learning Efficiency group at Meta Probability, focused on scaling AI across the organization. We developed a number of new algorithms: (1) A dynamic clustering algorithm for embedding tables, reducing the size of the internal recommendation systems by 50%. (2) A fast inference algorithm for Bayesian Neural Networks based on sketching. (3) Lower memory and latency transformer architectures for Facebook Assistant, enabling mobile deployment of language models. -
Postdoctoral ResearcherIt-Universitetet I København Feb 2019 - Mar 2020Copenhagen S, DkResearcher at the Basic Algorithms Research Copenhagen (BARC), a leading center for fundamental algorithmic research, headed by Mikkel Thorup. I developed a number of new algorithms, publishing papers at the top algorithmic conferences, SODA and FOCS. Some selected results: (1) An optimal LSH data structure for set similarity search. (2) A very fast sketching based algorithm for high-degree polynomial machine learning kernels. (3) Methods for doing similarity search with tensor cores. -
Phd CandidateIt-Universitetet I København Feb 2015 - Feb 2019Copenhagen S, DkMy research focuses on making machine learning efficient both theoretically and practically. The algorithms are often simple to use, even though the analysis may be mathematically complex. A secondary focus is making algorithms that are more robust, and on finding the limits of modern computing.
-
Co-Founder And Chief Machine Learning OfficerSupwiz Sep 2017 - Oct 2018Copenhagen, Copenhagen, DkI lead a team developing our chatbot software and putting it into production at three of Denmark’s largest IT firms. (There are now many more.) We used a combination of traditional symbolic AI and modern (at the time) sentence embeddings. The chatbot was awarded the most prestigious prize by the Innovation Fund Denmark in 2019. I was also in charge of our hiring efforts, interviewing dozens of candidates and hiring four engineers over the course of five months. -
Software EngineerSophion Bioscience A/S Oct 2013 - Feb 2015Ballerup, Capital Region Of Denmark, DkSophion Bioscience is a leading biotech company that specializes in developing advanced solutions for ion channel research. During my time at Sophion Bioscience, I worked on developing internal debugging tools that were used to sift through gigabytes of data per second on ion channel screening machines. -
OwnerPychess 2006 - 2015Developed the most used chess client and engine for the Linux desktop. Currently the 7th most used interface on the Free Internet Chess Server. Translated to more than 25 languages. I lead a team of 4-8 developers and designers. In 2009 we won Les Trophées du Libre in Paris. The project is GPL2 and GPL3 and has been used by people all over the world for research projects and other experiments.
-
Software Engineer InternPalantir Technologies Aug 2012 - Oct 2012Denver, Colorado, UsAt Palantir Technologies, my primary responsibility was to port the Metropolis ontological time-series system (now known as Foundry) to the web, which involved a series of complex tasks that required innovative solutions and careful planning. Over the course of this project, I focused on designing data visualization tools, efficient processing pipelines, and improving the overall user experience. -
Android And Php DeveloperXion Jan 2010 - Feb 2012Copenhagen K, DkI developed robust scrapers for 100 TV station websites, extracting valuable information such as emails, PDFs, and Word documents. Utilizing this data, I designed and developed a user-centric TV-listing app for Android, catering to over 50,000 users. The app's design was based on in-depth UI research and user statistics, ensuring an intuitive and engaging user experience. -
ConsultantFunch Aps Aug 2009 - Oct 2009I enhanced a large jQuery application's flexibility and performance by optimizing and refactoring the codebase. Additionally, I developed modules for XML and PDF conversion, enabling seamless data transformation and improved workflow. My expertise in UI design and development contributed to the project's success by delivering a more user-friendly and versatile tool.
-
DeveloperWine Doors 2006 - 2007Wine Doors was a project aimed at creating a package manager-like tool to facilitate the installation of Windows programs on Linux systems. As the first non-founder member of the project, my primary contributions involved developing user experience modules. Additionally, I focused on refining the codebase to align with Python standards, ensuring a clean and efficient development process.
-
Php DeveloperDaintel Jun 2006 - Aug 2006Kongens Lyngby, DkAt Daintel, a healthcare technology firm now acquired by Cambio Healthcare Systems, I worked on developing ICU and medical drug handling software. Using SQL & PHP, I created an optimized and reliable solution that enhanced patient care and supported healthcare professionals in their daily tasks.
Thomas Dybdahl Ahle Skills
Thomas Dybdahl Ahle Education Details
-
University Of OxfordComputer Science -
It-Universitetet I KøbenhavnComputer Science
-
Københavns Universitet - University Of CopenhagenComputer Science
-
Dansk Ungdoms FællesrådProject Development
-
Birkerød GymnasiumMathematics
-
Ecole Européenne De BruxellesMathematics
Frequently Asked Questions about Thomas Dybdahl Ahle
What company does Thomas Dybdahl Ahle work for?
Thomas Dybdahl Ahle works for Normal Computing
What is Thomas Dybdahl Ahle's role at the current company?
Thomas Dybdahl Ahle's current role is Head of ML.
What is Thomas Dybdahl Ahle's email address?
Thomas Dybdahl Ahle's email address is ahle@fb.com
What schools did Thomas Dybdahl Ahle attend?
Thomas Dybdahl Ahle attended University Of Oxford, It-Universitetet I København, Københavns Universitet - University Of Copenhagen, Dansk Ungdoms Fællesråd, Birkerød Gymnasium, Ecole Européenne De Bruxelles.
What are some of Thomas Dybdahl Ahle's interests?
Thomas Dybdahl Ahle has interest in Mathematics, Kayaking, Children, Scouting, Philosophy.
What skills is Thomas Dybdahl Ahle known for?
Thomas Dybdahl Ahle has skills like Php, Java, Python, Sql, Usability, Algorithms, Programming, Open Source, Software Development, Linux, Javascript, Community Management.
Free Chrome Extension
Find emails, phones & company data instantly
Aero Online
Your AI prospecting assistant
Select data to include:
0 records × $0.02 per record
Download 750 million emails and 100 million phone numbers
Access emails and phone numbers of over 750 million business users. Instantly download verified profiles using 20+ filters, including location, job title, company, function, and industry.
Start your free trial