Digital Data
Case Studies
AT&T pulls sales leads from multiple systems to automate data entry into legacy systems using Robotic Process Automation
Using RPA AT&T has been able to automate data entry. Data related to sales leads is extracted from various sources and is automatically filled into enterprise system freeing up hours worth of time for employees. They have also been able to improve customer service by reducing the time taken to offer certain services like ethernet from days to hou rs.
Linked use case: Improve administrative productivity with Robotic Process Automation
Adobe Research team investigates current approaches in machine learning for automating data cleansing and finds them inadequatee
Researchers from Adobe Research investigate the ability of metric learning techniques to automatically clean data. These approaches assume that datasets can be described in pre-defined ways, and that cleaning methods will work similarly well for similarly defined datasets. Ultimately, however, they find this not to be the case with the implication being that current cleaning standards are inadequate.
Linked use case: Automate data cleansing and validation
Anglia Ruskin University researchers develop mobile system which detects tuberculosis with 98.4% accuracy
Researchers from Anglia Ruskin University test different machine learning methods to classify digital images created using biosensors for the presence of tuberculosis antibodies. The goal was to create a system which could process and classify the images on a portable phone. They were able to achieve a 98.4% accuracy doing so using a random forest method.
Linked use case: Diagnose known diseases from scans, images, biopsies, audio and other data
Atos SE deploys Robotic Process Automation to automate extracting information in IT tasks
Atos SE automates IT tasks such as ticket management, incident management and server load balancing. These tasks were previously done manually.
Linked use case: Improve administrative productivity with Robotic Process Automation
Aviva accelerates from 400 to 10 days post-merger organisational planning and employee data integration with the use of machine learning
Aviva wanted to accelerate the post merger integration with Friend''s Life. They used Concentra''s AI powered OrgVue to help design a new integrated organisation and merge employee data from disparate systems. Time was reduced from a typically 400 days to ten.
Linked use case: Optimise staff and resource planning
BlueDot identified Wuhan pneumonia outbreak from social media posts before WHO made public announcement on COVID-19
BlueDot picked up on a cluster of “unusual pneumonia” cases happening near a market in Wuhan, China, and flagged it. This would become better known as the epicentre of what would come to be known as COVID-19. This was nine days before the World Health Organization released its statement alerting people to the emergence of a novel coronavirus in China.
Linked use case: Track and predict disease vector in general population
Client post trade equity allocation automatically generated from client emails at UBS
Following completion of share trades the client will need to have the assets allocated across various funds. Typically they email this through to the UBS team who, historically, then manually allocated the relevant shares. This repetitive task has been replaced by machine reading and data capture on the emails.
Linked use case: Automate collection of banking data from non-standard documentation
Dataminr''s alerts its customers in real-time about events that could potentially impact the financial markets by using text mining to collect data from social media
Dataminr''s AI platform analyses hundreds of millions of tweets everyday and classifies them against 30 rules and uses a proprietary classification system to build connections among thousands of companies, tickers, brands, and subsidiaries. They are able to notify customers several minutes before the news breaks out.
Linked use case: Automate customer service disruption alerts
Dun & Bradstreet saves 10,000 hours of work annually by using natural language processing to categorise five million firms in its directory
Dun & Bradstreet provides detailed information about five million businesses in the UK. To correctly categories the businesses into 1,000 industry classifications took a team over 10,000 hours annually. Using natural language processing to scan the internet to learn the jargon of a category it for the most part automated the process of business categorisation.
Linked use case: Automatically classify businesses
Elections Canada plans to use AI to identify electoral misinformation campaigns on social media
Elections Canada plans to use AI to monitor conversations on social media with regards to elections. The intent is to identify misinformation campaigns before they spread as a safeguard.
Linked use case: Censor user generated content on social media and other platforms
Facebook create new location address schemes for developing areas decreasing travel times by 22% in tests
Facebook researchers develop a model of predicting roadways and other human infrastructure from satellite images in order to create an accurate addressing system, particularly for areas where no consistent system exists. The goal is to aid developing countries and improve direction accuracy to help disaster relief. The automatic addressing system covers more than 80% of populated areas.
Linked use case: Create maps from satellite and other remote imagery
Graphnet detects changes indicative of a seizure in epileptic patients using wearable tracker and machine learning and notifies the patient and carer via app
Graphnet''s wearable tracker is constantly monitoring vitals such as heart rate, skin conductance, gyroscope, and accelerometer as well as information related to mood, stress levels, alcohol intake, diet and medication compliance. The machine learning system can identify patterns that lead to high risk of seizures to notify patient and care provider in time.
Linked use case: Alerting and diagnostics from real time patient data
Indix automates and standardises product information retrieved from a variety of sources using machine learning
Indix uses machine learning techniques to sort through product data collected from non-standardised and disparate sources for inclusion in its product catalogue.
Linked use case: Collate product information from multiple sources for central access
Instagram automates image description generation using machine learning
Instagram has announced it is introducing an alt-text generating tool which will provide visually impaired users with text or audio descriptions based of Instagram photos. This will evidently require some form of image classification.
Linked use case: Assist visually impaired people by describing images or reading text
Instapaper reviews saved articles and generates clean text ready to be read through their app using natural language processing
Instapaper uses Diffbot to curate content from web articles and newspapers and saves metadata to auto-sync with the app. Diffbot uses NLP and machine learning to extract content from web pages.
Linked use case: Leverage a Natural Language Processing library-SDK-API to support developers
Japan to automate patient information entry and some medical procedures by investing in 10 AI-powered hospitals
The Japanese government plans to invest more than $100 million in over half a decade to develop 10 hospitals that will be enhanced by AI, by 2022. AI-assisted programs will be able to complete patients'' charts according to conversations with doctors during examinations, read magnetic resonance imaging and endoscopic imaging and even analyse blood samples. The initiative aims to enable short-handed doctors to focus on patient care, combat chronic lack of doctors and nurses in some areas and also cut down medical expenses.
Linked use case: Alerting and diagnostics from real time patient data
Kik, a popular messaging platform, uses machine learning to integrate and scale different systems
Kik, a popular messaging platform, uses Databricks'' AI platform to integrate and scale different systems as it was having difficulty keeping pace with the exponential growth in user base and contents.
Linked use case: Scale and support data management and monitoring
LeasePlan, a vehicle management group in the UK, simulates different models to see cost distribution across various dimensions to decide which changes are optimal
LeasePlan, a vehicle management group in the UK uses OrgVue to design more efficient operating models. Using OrgVue, they can simulate various models to see the impact of change and choose the most cost-effective one.
Linked use case: Accelerate data integration from multiple sources
London Metropolitan Police plan automatically detecting illegal and inappropriate material on confiscated devices
The London Metropolitan Police plan to implement AI software which can identify inappropriate and illegal content, such as child pornography, stored on confiscated devices. Currently their software identifies guns and drugs.
Linked use case: Identify illegal or inappropriate images on devices confiscated by the police
Maria Fareri Children''s Hospital deploys a diagnosis platform using NLP and algorithmic search to assist physicians with difficult to diagnose symptoms
A diagnostic assistant platform called Isabel uses natural language processing and machine learning to offer possible diagnoses for physicians. Symptoms are entered into Isabel as text and the system''s algorithms search through its medical database to offer likeliness of different diagnostic predictions, with claimed 95% accuracy.
Linked use case: Assisted or automated diagnosis and prescription
Metlife reduces document processing time by 40% by automating data capture from manual documents
Metlife uses Captricity’s READ engine to decipher data from manual documents. The READ engine leverages computer vision to convert hand written material into digital format, processing upto a million documents in an hour. Metlife has been able to automate 42% of documentation capture process. The engine handles exception such as unreadable characters by tagging them for human review.
NASA automates 86% of HR transactions through Robotic Process Automation, RPA, of administrative tasks
NASA implemented Robotic Process Automation (RPA) on four administrative tasks in IT, accounts and HR in a shared services environment. RPA based business-process automation replicates typical back office tasks involving data manipulation and entry between multiple systems such as updating a customer address, for example. NASA claims 86% of transactions were completed without human intervention in their HR processes.
Linked use case: Digitise and automate processes using Robotic Process Automation (RPA)
Pharmaceutical company identifies warnings for non-Hodgkin’s lymphoma patients requiring change of treatment using machine learning
An undisclosed Pharmaceutical company has leveraged artificial intelligence to better understand non-Hodgkin’s lymphoma''s clinical progression and identify the best personalised treatment for each stage. The company applied machine learning to electronic health records (EHR) and other data to map out the warnings that indicate that patients need to switch to a later line of therapy. The team used an automated-feature-discovery (AFD) machine learning engine to test million hypothesis based on internal and external data. The system tried to find a statistically significant correlation between variables in patient data and transition to a later line of therapy. The technology enabled them to identify and isolate the variable combinations that predict transitions.
Linked use case: Predict personalised health outcomes to recommend individual treatment approach
Primer AI improves coverage of women of science using natural language processing
Primer AI''s Quicksilver system was fed 30,000 English Wikipedia articles on scientists, over 3 million sentences from news documents describing them and their work and names and affiliations of 200,000 authors of scientific papers. It discovered that there are 40,000 people missing from Wikipedia even though they have a similar distribution of news coverage as those who do have articles. It also discovered that women were underrepresented and has seen been used in three English Wikipedia editathons to improve women scientists''s coverage.
Linked use case: Update and improve corporate Intranet content
Researchers from various universities and labs enhance flood data collection with natural language processing and machine vision
Researchers at UC Berkeley, University of Dundee, Oak Ridge National Laboratory, Tufts University and Blue Urchin LLC have employed social media and crowdsourcing data to address the lack of hyper-resolution datasets for urban flooding. Data gathered from Twitter and a crowdsourcing app, MyCoast, are analysed using natural language processing and computer vision. The complementary data gathering method can improve detailed flooding risk analysis, urban flooding control, and the validation of hyper-resolution numerical models.
Linked use case: Assess building damage from weather or natural disasters with drones and satellite delivered imagery