Jay Buckingham Email & Phone Number

Edinburgh, Gb

• Led the machine learning team at Zelim, focusing on improving sea rescues in offshore energy sectors.• Developed Deep Learning models to locate people at sea during rescue missions using video feeds from drones, aircraft, and ships. Extensive use of Nvidia GPUs for model training and inference.• The success of these models helped Zelim win its first two customers with over £2M ARR.

Vancouver, Washington, Us

Developed deep learning models (neural networks) to help predict which auxiliary components to optimize the expression of particular genes on e. coli plasmids. For example, which chromosome chaperones work best to help create particular target proteins.

San Francisco, Ca, Us

Using machine learning to build models that help people find clothes they'll love.

Redwood City, California, Us

• Developed neural network algorithms for drones to detect structural defects in civil infrastructure (semantic segmentation via U-Net).• Conducted experiments to create models for detecting cracks in concrete and mechanical flaws.• Contributed to the advancement of drone technology for infrastructure inspection.

Helped our clients make use of machine learning techniques to better serve their customers. Focusing on Deep Learning for classification and recommendation tasks.

Zignal Labs delivers insights across the full range of social media, blogs and print news. They enable their customers to quickly spot trends, see relevant stories unfold, and take part in the conversations. Their data science team uses a range of machine learning and natural language processing techniques to detect sentiment and topics in articles, highlight the most influential people in conversations, and predict which stories will be truly important.

Cupertino, California, Us

Deep learning for object recognition. I explored a range of convolutional neural network architectures for recognizing objects in scenes. Primarily used Caffe, TensorFlow and OpenCV via Python and C++ with CUDA on linux machines with Nvidia GPUs.

San Francisco, California, Us

Hearsay Social is the leader in social business for the financial services industry, helping advisors and agents efficiently and successfully use social media to attract prospects, retain customers, and grow business. I led a small team of data scientists as we applied machine learning techniques to social media data to help advisors better understand their customers and their needs. We focused on using supervised learning algorithms to build models that detect important events in people lives, and unsupervised learning algorithms such as LDA to extract a person's key interests from their prose.

Dynamic Signal is helping brands connect with people who are interested in their products. My role is to develop algorithms to find those people. For example, I have used machine learning algorithms to create classifiers to determine the kinds of topics people talk about in their blogs and to determine the demographics of Twitter users.

San Francisco, Us

I was a research engineering lead at Klout. My team was basically doing "PageRank for people" -- computing the influence that people have on their online friends in particular topics. I also used machine learning algorithms to determine demographics relevant to advertisers for campaign targeting.

Redmond, Washington, Us

I was the program manager on Bing Relevance and Revenue team focused on helping Bing serve ads that users really want to see. My team used machine learning techniques to create the click prediction algorithms that predict whether a user will click on a particular ad. These algorithms are critical to maximizing Bing's ad revenue.

Us

iComprehend developed highly accurate image classification software for tasks such as content filtering and image search. I architected our flagship product, iClassify, and developed the image classification algorithms. The product was best-in-class with better than 95% accuracy in our core application.

Redmond, Washington, Us

Researched new machine learning techniques for detecting spam that greatly increased the amount of spam detected for Microsoft email products such as Hotmail, Outlook and Exchange. I received a Microsoft Innovation Award and was first author on several patents for this work.

Rulespace's core product was a service to filter inappropriate websites (pornography, hate, violence, etc.) I was one of the principal research engineers. I used machine learning algorithms to create highly accurate classifiers to categorize web sites. This was acquired by Microsoft in 2002 for parental controls at MSN.

Developed a wide range of pattern recognition, image classification, industrial robotics and embedded systems software for technology clients such as HP and Xerox. For example, I developed:- Computer vision software to inspect parts of inkjet cartridges using machine learning techniques to detect defects.- Embedded object tracking software for an aircraft-mounted infrared imaging system.