Deep Tech

SITUATION: John Lewis Instagram team 2mon pilot with Link Big

TASK: Get more IG performance, getting more site traffic from IG to JL site.

ACTION: Added ML codes to our cloud platform that track IG texts (IG API) and the engagement rate on JL IG.

RESULTS: Found out the cold words ("This Shirt") get more CTR than hot words ("this amazing jacket") - gaining 26% more organic traffic in 1 month!

SITUATION: Udacity's Nanodegree "AI for Product Managers".

TASK: Teaching ML to track the right signals for Pneumonia in Children X-rays images.

ACTION: Using APPEN, ML codes were trained to "learn" what are the 2 signals of Pneumonia in the lungs X-ray and flag which ones are "clear" or not.

RESULT: doctors precious time identifying cases of Pneumonia In Children.

SITUATION: Google Neuroscience Hackathon.

TASK: How to make use of "brain signals" analytics - for commercial use.

ACTION: Sync brain signals from site visitors and "clear the clutter" of data and gain insightful results.

RESULT: Testing a shoe store, we managed to see a pattern when a user got happier when he saw men shoes and not women shows.

SITUATION: IBM Watson Hackathon.

TASK: To solve the 95% fail of Startup by "matchmaking" the right startups to the right VC.

ACTION: Using IBM Watson "auto-matchmaking", and Crunchbase API, we've design an interactive dashboard helping map which are the best startups for VC goals and the reverse, best VC for startup goals.

RESULT: 1 judge favorite hack.

SITUATION: Global Hackathon to fight Covid-19 (Healthcare).

TASK: How do we track Covid patients - BEFORE they tested positive?.

ACTION: Google Home will detect if the user is coughing or sneezing and alert local authorities and guide the user to seek medical aid.

RESULT: Team managed to get Google voice trigger after 3 achoo and also detect the difference between a cough sound and a sneeze sound!