Artificial Intelligence

Innovation Report

2018 Powered by
 Artificial Intelligence

Innovation Report 2018

Introduction
Over the past few years, deep learning has gone from an esoteric branch of AI, focusing on theory, to
being so mainstream that even an actress from Twilight has published an academic paper on it. This rise
of deep learning over traditional machine learning methods has also led to the creation of a host of new
platforms designed to help businesses improve work and workflow.

The idea of creating a machine that can ‘think’ in a way similar to the human brain is as old as modern
computing. In 1950, in his seminal paper ‘Computing Machinery and Intelligence’, computing pioneer Alan
Turing laid out several criteria to assess whether a machine could be said intelligent, which has since
become known as the ‘Turing test’. Thirty-six years later, computer scientist and cognitive psychologist
Geoff Hinton demonstrated how an artificial neural network with several layers could be trained to learn
nonlinear functions. In this model, each succeeding layer in the network ‘learns’ from the previous layer.
In 2006, Hinton devised a way to train similar networks to pass along only information related to specified
parameters, adding and training new layers until a deep neural network was created.

But it was the more recent appearance of large, high-quality labelled datasets, distributed computing
and the applications of graphics processing unit (GPU) computing which has powered the recent,
rapid advancement of deep learning. It is now possible to create deep learning neural networks which
operate fast enough and accurately enough to have practical, real-world uses. Because of this, we are
experiencing a paradigm shift in computing, an AI boom in which companies are spending billions to
develop deep learning AI technology. Springwise has been following this shift since its beginning, and we
have continually highlighted businesses at the forefront of the deep learning revolution.

At Springwise, we have found that, while there are a huge number of businesses that could benefit from
deep learning, in general, for deep learning to offer a practical business solution, a company must have
a need for finding complex relationships in large amounts of data, and a recurring need for predicting
things that either cut costs or create value. Some of the most relevant deep learning-powered business
transformations we have found involve improved targeting of sales and marketing, better informed
decision-making, increased productivity and automation of retail.

Powered by
 Artificial Intelligence

Innovation Report 2018

Introduction
Targeted sales and marketing

AI modelling and simulation techniques are already being used to help drive sales in a number of ways.
Platforms providing real-time data gathering, forecasting, and trend analysis can offer greater insight into
buyer behaviour. Artificial intelligence used in conjunction with CRM systems can automate functions
such as contact management, data recording and analyses, and lead ranking, while AI-enabled buyer
modelling can provide a prediction of a customer’s lifetime value. Recommendation systems are also used
to learn content preferences and push content that fit those preferences, allowing more targeted sales.

While sales may appear to be primarily about relationships, in reality it is one of the most data driven areas
of business. From names and contact numbers, to information on customer behaviour and purchasing,
sales depends on the ability to find and identify the best sales leads, and the leads most likely to be
converted. Traditionally, this information was collected by individual salespeople, but deep learning has
allowed AI software to take over the time-consuming task of identifying contact information and lead
scoring. Start-up Radius has spent years building a network-effect-driven data source called The Network
of Record. The Network is built using data collected from more than 1 billion anonymized and aggregated
business interactions. This data is then analysed by Radius’ AI software to provide intelligence to sales
teams which can help them directly target only the most qualified sales prospects for any product.

Local pricing knowledge is a sales challenge faced by companies that operate in multiple markets around
the world. Local businesses can adjust prices based on local knowledge, but this is more difficult for
businesses that operate globally, or whose customers are primarily online. In online shops, businesses tend
to offer the same prices to every visitor, adjusted for local currency – regardless of the competitiveness of
their prices in different regions. While the sales team may be setting prices based on those of other online
shops, customers may be comparing them to local alternatives the sales team may not even know about.
Darwin Geo-Pricing has developed one solution to this, using a deep learning algorithm to retrieve the
online shopper’s location and combine this with data mining to adjust the prices each customer is offered.
Darwin’s machine learning system continuously performs exploratory pricing, offering different visitors
higher or lower prices than others in the same location. By comparing the sales performance at different
price levels, Darwin can determine the optimal price level to pitch at each customer.

Powered by
 Artificial Intelligence

Innovation Report 2018

Introduction
Another sales challenge is how to cut through the advertising noise and clutter most customers are faced
with daily in order to deliver more personal and relevant content or products to each consumer. The
Weather Company, which was purchased last year by IBM and rebranded as IBM Watson Advertising, is
working to cut through this background ad ‘chatter’ with the first-ever consumer use of IBM Watson AI
technology for advertising purposes. Watson Ads uses natural language processing and machine learning
to listen, think and respond to consumers’ questions with personalized answers in real time. Each Watson
ad is able to understand the consumer’s questions and respond in ways that sound like natural, human
language. Consumers clicking on a Watson ad for, say, a Toyota Prius Prime, can then interact with the
ad to receive individualized answers to questions about the brand, other Toyota brands and automobile
shopping. The program can also steer the conversation in a natural way, similar to how real people interact.

Another approach to advertising made possible by machine learning algorithms is predictive advertising.
Predictive advertising is a subset of predictive analytics, which uses machine learning to predict future
outcomes based on behavioural patterns in historical data. In ad tech, predictive analytics is used to help
businesses target audience segments based on behavioural signals, personalise ads to be more relevant
to individual users, or optimise ad bids based on user data. There are a number of companies vying to
develop platforms for predictive advertising, including Programmai, which uses AI to identify and target
new prospects with ads based on what is already known about a company’s existing customers or visitors.
The company’s software can notify a business of how likely someone is to make a purchase, calculate
future lifetime value, and automate ad bidding strategy decisions – allowing companies to focus ads on
where they will have the biggest impact.

A similar approach is taken by Indianapolis-based Demandjump, which uses analytics to help marketers
decide where to place ads in order to take advantage of up-to-the-minute trends. Their artificial intelligence
marketing platform, dubbed TrafficCloud, collects data on page views, and can link customer activity
across devices in order to present a detailed analysis of traffic between sources. The company’s analytics
can then pinpoint which sites have the greatest influence in a brand’s competitive area, and which sites
can help drive the highest number of customers to the brand. DemandJump’s proprietary algorithms
use machine learning, graph theory, algebraic topology, and natural language processing, to “map each
client’s entire digital ecosystem” and show marketers exactly what to do next to maximize revenue growth
and stay ahead of market changes.

Powered by
 Artificial Intelligence

Innovation Report 2018

Introduction
Better decision making

Business decision making is a critical process, at all levels. Many business experts suggest that the future
of AI in executive decision-making actually lies in a partnership, with humans defining the questions to
be asked, or problems to be tackled, and having a final say on the best answer for their business, while
AI is used to analyse large volumes of data to provide a basis for the decision. For an example of why
humans will not disappear from the decision-making process, we need look no further than Uber. When
the companies’ AI pricing algorithm doubled ride prices in London during a major terror attack in 2017,
the company suffered from the resulting bad publicity and accusations of price gouging. The algorithm
is better at setting prices in real time than humans, but a human would have been more likely to realise
that doubling prices during a tragedy is not good business. This partnership approach can be seen in the
development of a number of recent AI decision-making platforms.

In Denmark, start-up Corti has developed an AI platform that listens in to phone calls to emergency
services to help detect signs of a heart attack, including signals in verbal communication, tone of voice
and breathing patterns. Corti compares the data from each call against millions of previous emergency
calls to find patterns indicating a possible heart attack. Corti can then alert the dispatcher to the possible
need for an ambulance. This early recognition is important because the chance of survival decreases by
around 10 percent for each minute treatment is delayed after a heart attack occurs, and because many
heart attack symptoms, such as shoulder pain, can be easily dismissed as a minor ailment.

Partnership in decision making was also envisioned by the makers of VALCRI (Visual Analytics for sense
making in Criminal Intelligence analysis), which is designed to aid the police and criminal investigators.
The system is the result of an EU-wide research project involving more than 15 organisations - including
law firms and behavioural scientists. VALCRI analyses criminal records and data sets related to social and
legal systems to suggest possible areas of inquiry for detectives and criminologists, possibly reducing the
footwork and human error involved in investigation. Because the AI algorithm learns more as more data is
added, it can also help identify patterns that could help head off crimes before they are committed.

Deep learning has even penetrated the world of professional sport, allowing coaches to make better
decisions based on AI-assisted analysis. Analysing a sports match involves tracking thousands of

Powered by
 Artificial Intelligence

Innovation Report 2018

Introduction
data points across hundreds of plays, and it is impossible for a coach to look at everything at once. At
Springwise, we have highlighted how Australian innovator Flixsense is developing a platform, currently
in Beta, which is able to track every aspect of football and cricket matches, such as passing patterns and
ball tracking. Flixsense’s AI learns to recognize people, action and objects from videos and can instantly
identify key events from a match, helping coaches to anticipate what move the competing team will likely
be making next.

Increased productivity

The ability of deep learning algorithms to rapidly analyse large amounts of data and identify patterns
has the potential to reshape how businesses respond to operational challenges and inefficiencies. One
of the ways in which AI is already leading to improved productivity is through automation of complex
tasks that require physical adaptability and agility. This type of work is perhaps most commonly found in
construction, which has always suffered from lagging productivity. One reason for this is that large projects
involve many different sub-teams, each installing different components. It is impossible for managers to
accurately track the tens of thousands of components or elements on such projects, and it can be weeks
or months before errors are discovered. In response, California-based start-up Doxel has developed an
artificial intelligence system that operates drones and robots to monitor every inch of a construction
project and alert managers of potential problems. Doxel’s deep learning algorithms can also measure the
quality and progress of the work in real time, comparing it to the original budget and schedule.

An Israeli start-up called 3DSignals is also working to improve productivity in heavy industry by using
AI guided wireless ultrasonic sensors to track changes in the sounds made by machinery. By analysing
the changing sounds of the machinery and comparing them to data sets about the sounds of normally-
working and malfunctioning machinery, the algorithm can spot potential breakdowns before they occur.
Being able to predict problems earlier can, in turn, allow operators to save time and money.

Productivity improvements can also be driven through better workflow management. According to a 2017
report by market research firm Research and Markets, the market for workflow management systems
is estimated to grow from $3.5 billion USD in 2016 to almost $9.9 billion USD in 2021 – a compound
growth rate of 21 percent. A major part of this growth is likely to be in AI-driven systems that can help
companies achieve productivity gains through greater automation of workflow tasks. One innovator

Powered by
 Artificial Intelligence

Innovation Report 2018

Introduction
in this area is AskPorter, an artificial intelligence property management platform currently in Beta. The
AskPorter platform can serve as both a digital personal assistant for property managers and a property
concierge for tenants. The ‘porter’ allows automation of property management duties such as production
and management of legal and compliance documentation and reports. The system can also respond to
tenant queries, using a natural language processing algorithm, and resolve many low priority questions
and requests from tenants, freeing managers to respond to more complex and urgent tasks.

Natural language processing is also being used to improve productivity in human resources through
automation of recruitment tasks. In large companies in particular, recruitment is a very labor intensive
operation, but the early stages of the recruitment process, including verifying qualifications and answering
initial questions from applicants, lend themselves to automation. Job search company Mya Systems,
which works with a large number of tech companies and start-up, has recently introduced a recruitment
automation bot dubbed Mya. Mya uses artificial intelligence and natural language processing to ask
and answer questions and screen job applicants, as well as schedule interviews and create candidate
shortlists. Mya can analyse candidates’ declared competencies against the job description to develop
contextual screening questions for each applicant. The platform uses a deep learning-based multiple intent
classifier, named entity extraction, and sentence semantic analysis to steer the screening conversation
and extract meaningful information from candidate’s responses. Based on candidates’ answers, the bot
can independently evaluate a candidates’ suitability for the position and determine how to proceed.
Automating the early stages of recruitment can help free recruiters and hiring managers to focus more on
interviews and closing offers.

Automating retail

From self-scanning to restocking robots, Springwise has highlighted many ways in which intelligent
automation is driving innovation in the retail sector. One company, Deepmagic, has recently developed
AI-enabled image recognition software which can analyse how customers act in stores, including what
they pick up, and what they put down. Deepmagic is also using the system to develop the Qick Store
Platform – a fully automated mini store, intended to be installed in apartment building lobbies, airports
and hotels. Customers would gain entry through swiping a card or via a facial scan, take the products they
want and leave. The image recognition system would note the items removed and charge the customers’
account automatically.

Powered by
 Artificial Intelligence

Innovation Report 2018

Introduction
Pricing is another retail area where we are seeing deep learning offer practical solutions. The ability to
change prices rapidly to reflect shelf-life, competition and market conditions can help retailers maximise
their profits at any given time. Rapidmathematix has developed a system that utilizes deep learning
algorithms and machine vision to dynamically price and promote products, to match their real-time value.
The company’s algorithm can constantly re-price items based on information it gathers on freshness,
location, inventory, customer need, and changes to costs and competitors’ prices. When combined
with electronic shelf labels, the system can instantly update prices based on constantly shifting market
conditions.

At Springwise, we have covered ways that augmented reality (AR) is increasingly used in business,
including in developing training manuals and allowing collaboration between engineers at different sites.
Now, AR is also being combined with AI to help steer customers towards the perfect product. RoomAR
Analytics is targeting retailers and manufacturers in the furniture and home furnishing industry with a
product that uses an augmented reality mobile app to visualize virtual products in the real world and offer
personalized product recommendations. End customers and sales staff can use the AR app to see what
products will look like in the customers home or office, using deep learning to analyse current trends, and
the customers’ existing furnishings and visualise those products that will best suit their living situation,
taste and style. The system will also offer personalized suggestions based on wishes and conditions of
customers.

The future is now

From improving productivity on large construction projects to helping customers choose the right
furniture, machine learning and deep learning technology are being applied to almost every aspect of
business and industry. The combination of big data and neural networks is helping to unlock the value of
data businesses already have by revealing patterns that they can use to create and improve offerings or
productivity, or to gain an advantage over the competition. The recent release of machine learning and
neural network model builders, such as Gluon and Google’s TensorFlow, are also making it easier for
software engineers to create and run deep learning systems without specialized training. This will make
deep learning tools increasingly accessible to even SMEs.

In the future, deep learning-based image recognition systems will be able to identify words on a screen
and read them aloud, help an autistic child decipher facial expressions, or provide instant translations of

Powered by
 Artificial Intelligence

Innovation Report 2018

Introduction
written or spoken words from one language to another. Deep learning may also eventually change our
relationship to work by allowing many routine jobs to be done automatically. It is estimated that using AIto
automate U.S. federal government tasks now performed by employees would save at least 96.7 million
working hours a year, at a cost savings of $3.3 billion USD. But that also means 96.7 million fewer hours
of work available for humans.

For businesses, deep-learning tools hold out the promise of more efficient, more productive and more
streamlined operations, and better-informed decision making. While some types of jobs may disappear,
new ones will be created, and these will allow people to focus on what they do best, just as being creative,
rather than waste hours in tedious or repetitive tasks. Ultimately, cognitive technologies like deep learning
will fundamentally change how many of us work—and these changes will likely come much sooner than
many think. This makes it vitally important for businesses strategic planning to evolve beyond a focus
solely on people to consider the interplay of talent, technology, and design.

SPRINGWISE EDITORIAL TEAM

Powered by
 At a Glance
WHAT
Targeted at B2B marketing,
operations, sales and analytics
professionals, this platform offers
up-to-date data and intelligence.

WHO
Radius

WHERE
United States

CONTACT
www.radius.com
[email protected]

B2B marketing platform provides AI-

driven data insights
Targeted at B2B marketing, operations, sales and analytics professionals, this
platform offers up-to-date data and intelligence.

We have previously written about marketing innovations that provide data insights to improve businesses.
For example, this company uses AI and data science to measure TV audiences. Another example is
an intelligence system that provides retailers with actionable data. Accurate data insights can help
businesses to create successful strategies, operate efficiently, drive revenue growth, and more. Radius,
a B2B marketing platform, is using predictive AI to provide businesses with actionable insights and a
competitive advantage.

Radius has a constantly improving data set, called The Network of Record. The company currently provides
four different products to businesses: Radius Revenue Platform, Radius Advertiser, Radius Prospector, and
Radius Connect. The Revenue Platform uses predictive intelligence to drive revenue. Radius Advertiser is
a B2B advertising solution that connects with buyers on over 500 channels. Radius Prospector accelerates
sales using data from The Network of Record. Radius Connect can quickly source company and contact
data and it allows users to add this information to Salesforce.

Furthermore, using the Radius Omnichannel, B2B marketers can initiate campaigns on a variety of channels.
This includes email, field, search and digital. Additionally, campaigns can be initiated on social channels
that were previously only available for use with B2C customers, such as Twitter, Google, Facebook, and
more.

The solutions provided by Radius improve a business’s marketing, advertising and sales operations. It
does so through promoting efficiency, productivity and accuracy using big data. How else can AI data
insights enable businesses to execute better strategies?

10

Powered by
 At a Glance
WHAT
A new pricing solution enables
e-commerce businesses
to personalise prices and
promotions based on a
customer’s location.
WHO
Darwin Pricing LLC

WHERE
Switzerland

CONTACT
www.darwinpricing.com
[email protected]

Geo-targeted pricing service optimises

e-commerce sales
A new pricing solution enables e-commerce businesses to personalise prices and
promotions based on a customer’s location.

Darwin Geo-Pricing is a new pricing service that generates geo-targeted prices and promotions. Created
by Darwin Pricing LLC, the service uses Machine Learning Artificial Intelligence (AI) to optimise the sales
of e-commerce retailers. Darwin Pricing uses artificial neural networks to provide a real-time model of the
varied prices in different locations.

Using Machine Learning algorithms, the service creates optimal prices and promotions for each customer,
depending on their geographical location. This enables e-commerce retailers and brands to personalise
their pricing for each customer, therefore generating more profit. Additionally, the real-time collection of
data responds to any market changes, further boosting the accuracy of the pricing system.

Through a number of strategies using the data provided by Darwin Pricing, businesses can boost their
sales. For example, constructing sales campaigns based on competing local retailers in any city, and
creating personalised discounts for customers about the exit without a purchase. Darwin Pricing also
offers a customisable coupon box, compatible with smartphones and tablets. This feature can grow social
media followers and increase sign ups to newsletters.

E-commerce businesses can install Darwin Pricing as a plug-in. The service is fully customisable so that
each business can tailor it to their specific needs. Moreover, Darwin Pricing uses REST API, an API that
uses HTTP requests to manage data. Therefore, along with standard e-commerce platforms, any online
store can incorporate it. Many plug-ins are already integrated in Darwin Pricing. For example, Shopware,
Shopify, Magento Commerce, osCommerce, Websphere Commerce, WordPress, PrestaShop, Hybris,
3dcart, Oracle Commerce and Demandware. If desired, other plug-ins are also integratable.

We have previously written about Geo-spatial retail innovations such as a shopping app that alerts users
when an item on their wish list is nearby. Another example is an app used by customers to alert stores
about empty shelves. Darwin Pricing also enables businesses to increase profits. The service is unique
because it uses geographical data and personalisation to create more equal wealth distribution. How else
can e-commerce retailers and brands improve their performance using AI?

11

Powered by
 At a Glance
WHAT
Watson Ads enable marketers to
interact with customers in natural
language, as they gather data
about consumer behavior.

WHO
Watson Ads

WHERE
United States

CONTACT
www.ibm.com/watson
www.ibm.com/contact

AI for marketers to leverage cognitive

advertising
Watson Ads enable marketers to interact with customers in natural language, as
they gather data about consumer behavior.

In an increasingly noisy and cluttered world, advertisers are seeking ways of using technology to deliver
more personal and relevant content to consumers in ways that can scale to reach millions. Artificial
intelligence is leading the way in providing a solution. A Canadian startup created chatbots that use
AI to have conversations with users, and in Japan, a creative agency has appointed an AI robot that
mines creative databases to provide input as their creative director. The latest innovation comes from The
Weather Company in the form of Watson Ads, which use AI to answer consumer questions through the
medium of advertising.

The Weather Company, an IBM business since January this year, offers highly accurate weather data
to consumers and businesses. Watson Ads is their latest product and uses what is called ‘cognitive
advertising’ to interact with consumers. Users ask questions in natural language, as they would to Siri,
via voice or text and Watson can respond accurately. In their most recent collaboration with Macy’s, for
example, buyers can use Watson as a shop assistant to find out where certain products are in-store
or which brands they have in stock. Aimed at brands looking to have more valuable interactions with
customers, Campbell Soup Company, Unilever and GSK Consumer Healthcare were among the first to
collaborate with Watson.

What's more, even as consumers are learning about the product, Watson is gathering data from their
interactions in order to help marketers better understand brand perception, improve consumer experience
and inform creative strategies. Jeremy Steinberg, global head of sales at The Weather Company said of
the innovation, “We’ve embraced big data and leveraged it to improve every aspect of our business,
from forecast accuracy to ad targeting.” Are there more ways businesses can use this new technology to
improve consumer experience?

12

Powered by
 At a Glance
WHAT
A new advertising solution uses
machine learning to predict
customer behaviour and improve
advertising campaigns.

WHO
Programmai

WHERE
United Kingdom

CONTACT
www.programmai.com
www.programmai.com/contact-us

Predictive advertising identifies target

audience
A new advertising solution uses machine learning to predict customer behaviour
and improve advertising campaigns.

Programmai is a machine learning solution for brands and retailers to use for advertising. Using Artificial
Intelligence (AI), the tool can create predictive advertising campaigns.

Programmatic advertising is a way for businesses to target their campaigns at a specific audience. A more
accurate customer reach increases a business’s advertising impact. It also removes ad spend on those
that aren’t a part of the target audience. Programmai does this using machine learning methods. The
data collected by Programmai provides insights into the target customer’s traits and behaviours that were
previously unknown.

The company’s mission is “To disrupt how advertising is bought by producing ‘predictions’ that enable
you to bid smarter whilst anticipating your customer’s next move”. Programmai collects data that enables
businesses to make predictions such as how likely it is a consumer is going to make a purchase. The
technology can also help in the prediction of the lifetime value of a customer, measuring the entire future
potential a customer has.

Predictive solutions are changing the ways businesses operate. While Programmai is altering the
way businesses plan and measure advertising, other industries are also using predictive solutions to
streamline how a business operates.  A group of hospitals in France uses machine learning to predict
admissions. Another example is an app that can predict court decisions using big data. In what other ways
can predictive solutions enhance or re-structure how a business operates?

13

Powered by
 At a Glance
WHAT
An AI-driven platform aims to
predict exactly where and when
brands place their marketing for
the most impact.

WHO
DemandJump

WHERE
United States

CONTACT
www.demandjump.com
www.demandjump.com/company/
contact

New AI marketing tool provides detailed

analytics
An AI-driven platform aims to predict exactly where and when brands place their
marketing for the most impact.

How much more impactful could your marketing be if you knew precisely what content to write and where
to publish it to take advantage of up-to-the-minute trends? This is the question that Indianapolis-based
DemandJump hopes to answer. The company has developed the first Artificial Intelligence Marketing
(AIM) platform, dubbed TrafficCloud, which shows marketers exactly what to do next to maximize revenue
growth, outmanoeuvre the competition, and stay ahead of market changes.

While most marketing tools collect historic data, such as page hits, designed to help companies with
customer retention or cross-sell, TrafficCloud focuses on new customer acquisition. By using AI to collect
data for every event and page view, the company is able to link customer activity across devices and
present a much more detailed analysis of traffic between sources. The company’s analytics can then
pinpoint which sites have the greatest influence in a brand’s competitive area, and which sites can help
drive the highest number of customers to the brand. DemandJump claims that its proprietary algorithms,
using machine learning, graph theory, algebraic topology, and natural language processing, can “map
each client’s entire digital ecosystem and then cull the data to reveal revenue growth opportunities
brands didn’t even know existed”.

The marketing technology landscape has changed rapidly over the past several years. In 2011 there were
just 150 companies offering marketing solutions, and today there are more than 5,000. These include an
AI that matches brands with social media influencers and a system that produces real-time in-store data.
This growth means that marketers must pay much closer attention to constantly shifting market dynamics
in order to stay ahead of the competition. DemandJump currently has around 14 customers, but claims to
have a number of major deals with multi-billion dollar companies pending, and is growing quickly. Will the
type of AI marketing tool developed by DemandJump prove to be the way forward for marketing?

14

Powered by
 At a Glance
WHAT
Corti helps medical personnel
make critical decisions in the
heat of the moment on calls to
emergency services.

WHO
Corti

WHERE
Denmark

CONTACT
www.corti.ai
www.corti.ai/contact

AI assistant helps detect heart attacks on

emergency services calls
Corti helps medical personnel make critical decisions in the heat of the moment on
calls to emergency services.

Early recognition of cardiac arrest is vitally important as the chance of survival decreases about 10 percent
with each minute. In Denmark AI assistant Corti is listening in to phone calls to emergency services to help
detect signs of a heart attack. With Corti implemented, the dispatcher gets a digital assistant that listens in
on the conversation and helps to look for important signals in both verbal communication, as well as tone
of voice and breathing patterns, while also considering other metadata.

All the data provided during the emergency call is automatically analyzed by Corti and then compared
to the millions of emergency calls – which Corti has already analysed –to find important patterns. As
Corti's understanding of the incident increases, the assistant will try to predict the criticality of the patient's
situation based on symptom descriptions and the signals gathered from voice and audio. These insights
are delivered to the dispatcher as alerts and recommendations.

Artificial intelligence is seeping into every aspect of every day life, with recent uses including the system
that hunts the internet for signs of suicidal people to help prevent them coming to harm and an AI
politician helping answer the public’s questions in New Zealand. How could artificial intelligence help
your operation?

15

Powered by
 At a Glance
WHAT
VALCRI, an AI algorithm,
processes criminal databases and
other data sets, suggesting lines
of inquiry to aid police, and may
even be able to pre-empt crime.
WHO
VALCRI

WHERE
European Union / United Kingdom

CONTACT
www.valcri.org
[email protected]

EU police forces trialling AI detectives

VALCRI, an AI algorithm, processes criminal databases and other data sets,
suggesting lines of inquiry to aid police, and may even be able to pre-empt crime.

We’ve already seen how, in China, an image-recognition algorithm is aiding surveillance in real-time, and
now UK police will be using similar algorithmic technology to assist criminal analysts.

The system, called VALCRI (Visual Analytics for sense making in CRiminal Intelligence analysis) is the result
of an EU-wide research project fronted by Middlesex University, involving 18 organisations of various
expertise from law firms to teams of scientists, and aims to herald the age of ‘intelligence-led policing’.
VALCRI analyses anonimized criminal records, and pulls in other various data sets that offer insight into
social and legal factors, in order to suggest likely lines of inquiry for detectives, a process that is classically
highly labor-intensive for criminal analysts and subject to human error or biases.

VALCRI is not intended to solve crimes, but offers data and suggestions of possible leads in a graphics-
oriented approach across two large touchscreen consoles. The system features an AI algorithm that is
capable of learning as more data is added, so that its ability to see patterns only grows more robust as
data is added. As well as assisting ongoing investigations, VALCRI also claims to be able to pre-empt
crimes by spotting developing patterns that may have historically led up to incidents of crime, and will also
provide a vast dataset for criminal researchers that may be used to inform policy on crime prevention. A
pilot project is underway in the UK with the West Midlands Police Force and in Brussels, whereby VALCRI
has been fed with three years worth of local criminal data.

What ethical and legal frameworks would need to be put in place if we were to see the onset of 'Minority
Report' pre-emptive crime intervention?

16

Powered by
 At a Glance
WHAT
A new technology powered by
deep learning offers coaches
actionable insights by tracking
every aspect of soccer and cricket
games from videos.
WHO
Flixsense

WHERE
Australia

CONTACT
flixsense.com/index.html
[email protected]

Sports video intelligence platform

provides instant match footage analysis
A new technology powered by deep learning offers coaches actionable insights by
tracking every aspect of soccer and cricket games from videos.

We have already seen the effect of innovation on the sporting world, from new tech used to take cricket
games to a new level to a wearable patch that analyses athletes’ sweat. The Australian innovation
Flixsense gives the world’s coaches and teams the unique opportunity to access instant match footage
analysis. The company uses AI technology to provide coaches with the insight they need to win.

The company currently consists of a three-person team, and is looking to expand. The platform is able to
track every aspect of the football game, from goals and passing patterns to ball possession and player
tracking. There is also a platform specifically designed for cricket which does ball tracking, batting stroke
analysis and player tracking. The technology uses deep learning algorithms to provides unique insights
into match footage which can be used to help coaches improve their team’s results. The AI models are
able to expand their knowledge and become smarter the more videos are added. The startup is hoping
to expand their technology to suit more sports. Consumers are currently able to sign up to ask for a demo
of the platform and to receive product notifications.

This is an innovative use of AI technology which we have seen implicated in other parts of the fitness
industry, such as providing tailored workouts to gym goers and smart calorie counters, however Flixsense
provide a unique service specifically aimed at world standard team sports. How else could AI technology
be used in the professional sporting industry? How might this video platform revolutionise worldwide
coaching techniques?

17

Powered by
 At a Glance
WHAT
A new system uses drones and
AI to improve accuracy and
efficiency on large construction
projects

WHO
Doxel

WHERE
United States

CONTACT
www.doxel.ai
www.doxel.ai/#press

New AI system set to disrupt the

construction industry
A new system uses drones and AI to improve accuracy and efficiency on large
construction projects

The construction industry accounts for around USD 10 trillion annually – that’s about 13 percent of the
global GDP. But up to now, this sector has lagged behind in productivity gains. This is due largely to
the fact that large projects involve many different sub-teams, each installing different components. It is
impossible for managers to accurately track every aspect of such projects. It could be weeks or months
before they realise an error was made. According to the McKinsey Global Institute, most large projects
take 20 percent longer to finish than anticipated, and run up to 80 percent over budget. California-based
startup Doxel is hoping to change this with an artificial intelligence system that can monitor every inch of
a construction project and alert managers of potential cost overruns, inefficiencies and errors.

We have already seen the use of drones and robots to navigate spaces people can’t reach. Some
examples are a robot snake that can crawl through pipes and a robotic eel that can search out water
pollution. Doxel significantly expands this by using autonomous devices, such as drones and robots. They
are equipped with LIDAR and HD cameras to visually monitor an entire construction project, even in small,
dark spaces that are difficult for humans to reach and see. The system then uses deep learning algorithms
to measure the quality and progress of the work in real time, comparing it to the original budget and
schedule. The system can process three-dimensional data and determine how much material has been
installed correctly, and where the project is in danger of running over-budget. Project managers can then
use this information to make changes early in the project and boost productivity. Errors can be spotted
and fixed immediately, before they lead to cost and time over-runs.

Doxel is designed for use on large projects with contracts of USD 20 million and above. The system
provides project managers with a cloud-based dashboard which can keep track of the hundreds of
thousands of items in a project in real time. This could allow them to focus on making real improvements
based on accurate data. The company has recently raised USD 4.5 million from a venture capital group
led by Andreessen Horowitz. What other types of projects could benefit from a similar system for keeping
track of progress?

18

Powered by
 At a Glance
WHAT
A new recruitment chatbot helps
each candidate with their job
search while streamlining the
process for employers.

WHO
Mya Systems

WHERE
United States

CONTACT
hiremya.com
[email protected]

AI recruiter does all the tedious tasks

A new recruitment chatbot helps each candidate with their job search while
streamlining the process for employers.

Looking for a job is generally time-consuming and stressful. Candidates often hear back from only a
handful of recruiters, while employers stress that they aren’t finding the right people. Formerly known
as FirstJob, Mya Systems's Artificial Intelligence (AI) chatbot is hoping to improve the process for both
employers and jobseekers.

As soon as a candidate applies, Mya gets in touch via instant messaging. If she spots gaps in a resume,
she will ask detailed follow-up questions. Candidates then get the chance to better explain how they
fit the position. For employers, Mya ranks candidates according to qualifications, prioritizing those who
answer the most frequently asked questions about each company. Mya stays in touch with candidates
throughout the process. It send alerts about other potentially suitable jobs and when a position has been
filled. Mya also helps employers by handling time exhausting tasks such as phone screening and interview
scheduling. Targeting millennials, the service is free for jobseekers, while companies pay a fee based on
the number of applications they process each month.

We’re seeing a number of different ways AI is being used to assist employees, from cognitive advertising
to beer brewing that evolves via feedback. What industries have yet to try AI-based, smarter, and faster
solutions?

19

Powered by
 At a Glance
WHAT
An Israeli startup is using
deep learning and ultrasonic
technology that could
revolutionise the automotive
industry.
WHO
3DSignals

WHERE
Israel

CONTACT
www.3dsig.com
[email protected]

Ultrasonic sensors detect mechanical

problems before they occur
An Israeli startup is using deep learning and ultrasonic technology that could
revolutionise the automotive industry.

When it comes to machinery, diagnosing specific problems can be difficult. Now, a new Israeli startup
called 3DSignals is using wireless ultrasonic sensors to track changes in the sounds made by machinery,
to predict when breakdowns are about to occur.

The technology uses deep learning to listen for deviations from the norm in musical and mechanical
acoustics. It identifies anomalies, classifies patterns of equipment failure and predicts issues before they
interrupt production. This has a number of benefits. First of all, it reduces the impact of human error,
which can cause costly unplanned downtimes. Secondly, it means users can react to problems quicker,
meaning perishable components are less likely to be significantly damaged. The tool also allows machine
operators to keep track of the condition of their equipment and plan maintenance breaks accordingly.
3DSignals is all about increased efficiency and, according to the founders, "When trained, the 3DSignals
deep learning algorithms are able to identify [and] predict specific problems in advance with 98 percent
accuracy,"

Innovations in the automotive space continue to appear. In September last year, we covered a startup who
launched a fleet of driverless taxis. Given car giants like BMW and Volkswagen's aim to launch their own
self-driving taxis, rather than consumer vehicles, will the focus be on applications for autonomous taxis?

20

Powered by
 At a Glance
WHAT
With an intelligent assistant
interface, this new platform helps
tenants and managers handle
queries and routine contractual
work quickly and efficiently.
WHO
AskPorter

WHERE
United Kingdom

CONTACT
askporter.com
[email protected]

AI property management saves time and

money
With an intelligent assistant interface, this new platform helps tenants and
managers handle queries and routine contractual work quickly and efficiently.

AskPorter is a white label artificial intelligence property management platform. The company’s goal is to
help reduce inefficiencies in workflows, ultimately transforming the role of property manager into that of a
supervisor or concierge. Porter is the natural language processing assistant who handles all initial queries.

Designed to bring together the many different aspects of property management, the AskPorter interface
is the same for all users, whether a tenant, owner or manager. Accessible via mobile, desktop or email,
the platform seamlessly moves between devices. Property managers can see on one screen all similar
documents or all stakeholders involved in a specific property or set of properties. Customized reports
can be produced and shared, ranging in detail from a specific activity across all managed properties to all
relevant information for a single property.

The more the system is used, the smarter Porter becomes. By acting as the first point of contact for queries,
many of the low priority, easily resolvable yet time-consuming questions and requests from tenants can
be managed quickly and easily. When the assistant needs advice, the AI knows which team member to
bring into the conversation.

Regular tasks, including management of legal and compliance documentation, can be automated, freeing
up a manager’s time to focus on building relationships and creating more personalized solutions. The
AskPorter team believes that the platform could help reduce property management costs by up to 50
percent. Closed beta trials begin in December 2017, with a public beta version available at the end of
quarter one in 2018.

Natural language processing is an exciting area of exploration, with industries ranging from retail to
healthcare testing out the possibilities. For sellers, a system of virtual sales assistants helps improve
interaction with customers, and for understaffed medical facilities, a chatbot supports doctors by speeding
up the process of diagnosis and offering treatment options. How might partnerships between businesses
using artificial intelligence provide additional products and services?

21

Powered by
 At a Glance
WHAT
A new retail solution uses artificial
intelligence to enhance the
shopping experience and enables
stores to operate unattended.

WHO
DeepMagic

WHERE
United States

CONTACT
www.deepmagic.cc
[email protected]

Startup uses AI to fully automate retail

stores
A new retail solution uses artificial intelligence to enhance the shopping experience
and enables stores to operate unattended.

We have seen many retail innovations demonstrating that the future of retail is both digital and offline.
For example, a touchscreen corridor that uses image recognition and purchase histories to suggest
and locate items to customers. Another example is this smart shopping trolley that charges customers
automatically. A new startup, called DeepMagic, also uses digital innovations to enhance the experience
of a physical store. Using Deep Learning Artificial Intelligence (AI), DeepMagic is enabling retailers to
create unattended physical stores that are fully automated.

DeepMagic offers two products: the Qick Store Platform and the Qick Kiosk. The Qick Store Platform is a
solution that lets customers shop with mobile scan and pay. Compatible with iPhone and Android phones,
the shopper app also has integrated anti-theft controls. Using AI, computer vision and cameras, DeepMagic
tracks the activities of customers in store and recognises any suspicious behaviour. Additionally, store
managers can access this information in real-time.

The Qick Kiosk is a pop-up store that is fully automated. Image recognition software enables these
unattended mini-stores to operate securely. The ideal positioning of Qick Kiosk is in places such as hotel
lobbies and residential buildings – locations where shops do not exist but people would like to see them.
Customers sign up on the Qick portal or mobile app. After signing up, customers can access and shop in
the Qick Kiosk stores.

Products taken from retailers using DeepMagic are automatically charged to a customer’s account.
Alternatively, customers can QickScan a label to arrange a home delivery. Unattended stores provide
convenience to both retailers and customers through extending store hours, saving costs and automating
check-out and security. An increasing number of retail solutions are using AI to provide a more seamless
shopping experience for customers. Moreover, we are seeing that these solutions are not limited to a
single platform. Instead, omni-channel approaches are being used to draw in customers and increase
sales. What other AI solutions could facilitate retailers to digitally enhance the shopping experience of a
brick and mortar store?

22

Powered by
 At a Glance
WHAT
A new pricing system uses deep
learning algorithms and real-time
data to price produce and reduce
food waste.

WHO
RapidMathematix

WHERE
United States

CONTACT
www.rapidmathematix.com
[email protected]

AI algorithm predicts prices of produce

based on freshness
A new pricing system uses deep learning algorithms and real-time data to price
produce and reduce food waste.

We have seen many solutions that tackle the issue of food waste, such as grocery stores introducing new
pricing systems to ensure products close to expiring will sell. One example of this is a real-time pricing
solution that uses radio frequency identification (RFID), electronic shelf labelling, and a dynamic pricing
engine to offer cheaper prices. Another example is an automated discount rack that reduces prices for
expiring products both online and in store. A new solution, called RapidMathematix, also aims to reduce
food waste using deep learning algorithms and machine vision.

RapidMathematix provides automated retail pricing for fresh produce, changing the prices of produce
depending on freshness, market conditions and competition. This ensures that customers get their
money’s worth of what they pay and that stores are able to reduce food waste by offering discounts.
Data collected from various sources about freshness, location, product and demand level is processed by
RapidMathematix’s algorithms to offer the most accurate prices. Additionally, the system is connected to
electronic shelf labels, enabling it to calculate and recommend prices in real time.

IoT devices are also integrated into the RapidMathematix system and are used to gather information from
inside the retail store. For example, information is collected from products, the shop floor, shelves and
customer devices. The data offered by the system can also be used to negotiate prices with vendors, beat
competitor prices, and give users more control over their pricing decisions.

In the retail industry, promotions based on real-time data can reduce food waste and increase sales. What
other industries could increase sales by using deep learning AI to offer fast and targeted promotions?

23

Powered by
 At a Glance
WHAT
Using AR and AI, this app can be
used to visualise home furnishings
before purchasing products.

WHO
RoomAR

WHERE
Germany

CONTACT
www.roomar.de/home
[email protected]

Interior design app lets users visualise

products in the real world
Using AR and AI, this app can be used to visualise home furnishings before
purchasing products.

A new interior design app, called RoomAR, uses Augmented Reality (AR) and Artificial Intelligence (AI) to
enhance the home furnishing experience. Targeted at retailers and manufacturers, RoomAR allows users
to visualise how virtual products will appear in the real world.

Using Deep Learning, Machine Learning algorithms and computer vision, RoomAR Core analyses the
user’s sense of style. Taking style and living conditions into consideration, the mobile app can personalise
the product recommendations it offers to each user. After recommending products, the app connects to
an online shop through RoomAR Shop, enabling users to purchase products. A personal and easy-to-use
service, RoomAR offers an innovative way to furnish homes and enables businesses to increase sales.
Additionally, RoomAR is useable as a sales tool, to help advertise products at trade fairs and in showrooms.

Data-based business insights are also available to retailers and manufacturers through RoomAR Analytics.
Data providing insights into trends, customer activity and target customers, is collected using machine
learning techniques.

RoomAR is changing the way consumers shop for furnishing and the way businesses sell furnishing. By
offering AR visualisations of products, consumers and businesses can see exactly how a product will look
in an environment. RoomAR makes this possible by merging online and offline retail experiences together.
Here at Springwise, we have previously written about other businesses that also offer an omni-channel
experience. For example, this interior decorating app lets users experiment with room decor, while this
app uses visual recognition to search for furniture. What other industries could utilise AR and AI solutions
to increase sales and improve business insights?

24

Powered by
This report has been compiled by the Springwise Editorial team.
Powered by a network of over 21,000 Springspotters in 190+ countries,
Springwise curates and publishes the most exciting global innovations
every day. The Springwise database provides the most up to date
innovation intelligence in the world — to help you navigate the future and
create positive change. CEOs, Innovators and Leaders from around the
world are already gaining innovation advantage by tapping into a global
source of knowledge and insight that is helping them to fast-track product,
business and market innovations.

Stay ahead of the latest trends to win business, empower your teams,
reset your strategy with www.springwise.com

Questions?
[email protected]

Copyright © 2018
Springwise Intelligence Ltd.
All rights reserved.

Powered
Poweredby
by