Automation Artificial Intelligence Machine Learning Deep Learning The

Automation, Artificial Intelligence, Machine Learning, Deep Learning & The Kitchen Sink James Ewing INTELLIGENCE AI Practice

Definitions (Level Set) Automation & Robotic Process Automation Physical or Software robots that complete a set of tasks based on instructions and process Scheduled back ups of IT systems on a regular basis and notifications (success/fail) Basic Artificial Intelligence Simple algorithms and rules that give the impression of intelligent computing Enemy player in early arcade games Machine learning Systems that have the ability to learn without being explicitly programmed. This type of AI learns through training and the quality of analysis improves with experience Predictive failure, Netflix recommendations Deep learning Advanced AI, based on practice and feedback, rather than rules. Requires high quality data that is processed analysed in a similar way to the human brain. The ability for a computer to detect tumours in medical images. INTELLIGENCE AI Practice

Overall Growth Impact Of AI & Automation - UK +$814 bil 319 416 3206 Baseline AI Steady State Baseline Intelligent Automation Augmentation INTELLIGENCE Source : AI Is The Future of Growth – Accenture 2016 AI Practice AI could add an additional US$814 billion in 2035 to the UK’s economy—with growth rates increasing from 2. 5 percent to 3. 9 percent in 2035. The growth boost will result in approximately equal parts from the augmentation and intelligent automation channels. While the UK’s dominant service sector can adopt AI to fuel the productivity of knowledge workers, Britain’s strong pharmaceutical and aerospace industry could also capitalize on intelligent systems to optimize production.

Automation - The Productivity Problem The inefficiency of human knowledge workers Disparate, disconnected applications Organizational structures and skills based teams INTELLIGENCE AI Practice

The Evolution of Process Robotics & Automation Business Functions Teams or Departments Business Processes Business Systems INTELLIGENCE AI Practice

AI & ML So Why Is It So Hot Right Now… 4 Factors Coming Together Doing Things Humans Just Can’t… • Consumer AI (Siri Alexa) has alerted us to Possibilities • Breakthroughs in learning algorithms Big Data: • Big Data sources- Customer Experience. • Io. T – Generates vast amounts of data • Consumer Data – AI helps make sense of it The 3 rd Platform • Cloud computing makes it possible Source : IDC Oct 2017 INTELLIGENCE AI Practice Recommendation engines • Monitor thousands of sensors in real-time to predict failure Pattern detection in large datasets e. g. photo tagging & video analytics

What Defines AI & Automation Assisting Supporting human tasks or decisions but do not learn or adapt Automation Conclude routine tasks or decisions, but do not change or require new ways of completion. Augmenting Supporting or improving human decision making and learning over time. Autonomous Making decisions and learning without human intervention. INTELLIGENCE AI Practice

Characteristics Conversation & Voice Tone, Sentiment & Empathy Vision (Image & Video) Listening & Audio Reading & Understanding INTELLIGENCE AI Practice

What is the Impact on Business? Reporting Forecasting Reacting Predicting Managing Optimising INTELLIGENCE AI Practice

Some Core Concepts Realisation Implementation AI Principles Conclude tasks beyond human scale Allow every employee to be as good as your best Replacement technology/people Enhance technology/people Augment existing processes & systems Provide new interactions & insights quickly and at scale Create new products, services and businesses Learn, enrich and extend INTELLIGENCE AI Practice

Enabling Multi-Channel with AI and Automation Update to Existing Application Inbound Structured Document Call to Contact Center Customer Self Service Inbound Unstructured Document Inbound Natural Language Web Chat or Virtual Assistant INTELLIGENCE AI Practice 1. Structured information extracted and utilized for automated process 2. Standard OCR extraction of information from defined and identified fields 3. Agent uses web form to provide information to process 4. Customer uses web form to provide information 5. Intelligent OCR to extract and then translate information and learn over time 6. Using AI & ML to translate language, identify and extract information, intent and context 7. Utilise Natural Language for inbound and outbound interaction

Enabling Multi-Channel with AI and Automation Update to Existing Application Inbound Structured Document Robotic Process Automation Call to Contact Center Customer Self Service Intelligent Automation Inbound Unstructured Document Inbound Natural Language Web Chat or Virtual Assistant INTELLIGENCE AI Practice AI / ML / DL & Predictive

Define Your Path To Automation & AI Work out what AI means for your business 1 2 3 Prioritise your responses Get the right talent and culture, as well as technology Think about governance up front INTELLIGENCE AI Practice

Predictions for 2018 & Beyond “ AI will begin to answer ‘Why? ’ • Humans are not good at not knowing what AI is doing • & AI is not good at telling humans what its doing “ “ “ ” Personalised Dynamic Pricing • There will be a price on the page or shelf • & then there will be ‘your’ price Software companies will be embedding AI, its like water or air around us– Its not a category. It will be in most software whether we know it or not ” INTELLIGENCE AI Practice ” “ 2018 the implementation of ‘smart automation’ will deliver the most immediate results. So many businesses still rely on decades-old, legacy-driven manual processes which create bottlenecks in the digital world ” In 2018, business leaders will push to make the business run more efficiently and will turn to AI and machine learning for help. Companies will also turn to AI to help scale and do jobs instead of adding headcount ” “ High quality training data will become a coveted resource, like oil for the information age. The sharpest human minds in tech may even shift their attention from creating algorithms to feeding those algorithms the best data diet “ ” AI applications will make a big splash in enterprise support functions in 2018, as call center, finance and IT executives begin to move conversational AI, image recognition and autonomic applications from pilot mode into production ”