Thursday 7th November, 2019
The Seminars will take place from 12:15 – 13:00
Delegates will be able to attend one seminar at the event. No pre selection is required – delegates will be able to select which session they attend onsite.
Conference Centre – Auditorium (Main Conference room)
Understanding Your Customers Inside Out: Using Data Assets to Balance Customer Retention and Profitability in a Shrinking Market
Simon Kelly, Director – Business Intelligence Group, CACI
The balance between customer churn and profitability is never an easy one. In a highly competitive market, even the smallest change in rate plans can have a significant impact on a company’s overall revenue.
Learn how one of Britain’s largest telecommunication companies combines big data technology, complex prediction algorithms and visual analytics to understand customer behaviour and its impact on the company’s profits. What was once a guessing game is now a more exact science and a key enabler in strategic pricing decisions.
Conference Centre Room A
The Path to Enterprise AI: Tales from the Field
Enterprise AI is a target state where every business process is AI-augmented and every employee is an AI beneficiary. But is that really attainable? And, if so, what is the path to get there? In this talk, Dataiku will share learnings from the field, describing how companies of different sizes and across different sectors have begun this journey. Some are farther along than others, and by making the right decisions now and avoiding stumbling blocks, you can supercharge your quest to this AI-fuelled future.
Conference Centre Room B
Data Modelling for Big Data & Analytics Projects with Apache Cassandra
Patrick Callaghan, Business & Technology Strategist, DataStax
Big Data and Analytics initiatives depend on data to work. That data has to be modelled and managed well to get value out of it. However, the approach we take around data modelling can be influenced by the ways that we store data over time.
Understanding this in advance can make it easier to query our data, and also avoid problems caused by bad decisions or overlooking data modelling at the start.
This session will help you understand how to build a strategy and implement a data model on Apache Cassandra to support your Big Data projects.
Conference Centre Room C
Building Predictive Maintenance Solutions for High-Speed Hitachi Trains in the UK
Dr Anya Rumyantseva, Senior Data Scientist, Hitachi Vantara Limited
Philip Hewlett – Digital Transformation, Hitachi Rail Limited
Hitachi Rail is introducing a fleet of high-speed trains on the UK railways – one of the busiest rail networks in the world. These trains, modelled on the Japanese bullet train, cut journey times by accelerating faster than existing rolling stock. But what if a train breaks while in service causing delays and network-wide disruption? Each Hitachi Rail train is carrying thousands of sensors generating trillions of data points every year. Application of advanced analytics and machine learning on these data allows predicting equipment failures, reducing in-service downtime, and optimising operating schedules.
Hitachi Rail uses the Pentaho platform of Hitachi Vantara to bring these analytical use cases to revolutionise rail transportation reliability in the UK.
In this seminar, we will discuss:
- Type and amount of data collected from the high-speed Hitachi trains
- AI/ML use cases for digitalising rolling stock maintenance
- Methodology for facilitating collaboration between data engineers, data scientists and rail domain experts
- Social impact of the project
Conference Centre Room D
Bridging the AI Gap: Uniting Data Scientists and Data Engineers for Accelerated Analytics Insight Using Talend
Ben Saunders, Expert Solutions Engineer, Talend
Today, many Data Science teams are choosing to leverage the power of MLaaS (Machine Learning as a Service) platforms on Azure, AWS and Google Cloud for developing and operationalising their AI and ML pipelines for predictive analytics. There are huge advantages to working this way such as lower TCO, accelerated development and scalable cloud solutions. However, this migration to MLaaS platforms is not without its challenges. Firstly, Data Scientists still spend far too much of their valuable (and costly) time locating, preparing and cleansing data prior to modelling. Secondly, integrating new ML pipelines into existing applications (typically managed by different teams) can be difficult and often slow, meaning many organisations struggle to achieve the desired ROI from AI initiatives. This has led some companies to build entirely new (often segmented) teams to manage the operationalising and deployment of their ML pipelines using a process called MLOps. If only solutions existed to make the hand off between Data Scientists and Data Engineers more seamless leverage existing processes and practices…
In this session we demonstrate how Talend can be used to dramatically enhance the productivity of your Data Science and Data Engineering teams by continually delivering high quality, trusted data to the Cloud whilst also enabling a viable MLOps solution.