Demonstrate real business value from deploying technology
Information Age brings together four of its most popular conferences – Data Leadership, Security Leadership, Mobile Leadership and Cloud Leadership – into one mega summit, attended by 600+ tech leaders.
SPEAKER: Toby Balfre - Head of Data Science, Capgemini UK
Toby is the Head of Data Science (DS) in the UK, for Capgemini’s Insights & Data Global Practice, working for clients on DS engagements in a range of sectors, particularly the public, utilities and consumer products industries.
He is active in the design and development of DS and analytical capabilities, including the agile development of DS proof of concepts and their transition to production services. Toby has in-depth experience with a range of Big Data and Data Science technologies, and has been championing the exploitation of Open Source technologies. He is, for example, proficient in both Python and R for Data Science, Graph databases using Neo4J, advanced text processing using ElasticSearch, and big data manipulation with SQL languages on Hadoop.
His team of 35 data scientists are currently working in the public, CPG and utilities sectors and he is overseeing the oil and gas, and water leakage capabilities whilst providing technical advice in the public sector. He has Masters Degrees in Applied Maths and Astrophysics.
Toby’s presentation at the Data Leadership conference will form part of the Technology Stream. He advises not to choose one database, but use them all to best effect. “Today there is a broad range of database technologies available on the market, optimised towards different types of data and different processing tasks. In every IT department you will find architects huddled in the corner sagely stating the right database for that company.
But why do we choose one database technology? As Data Scientists, we want to analyse the same data in many different ways to discover the most important and valuable insights. To do this we need the same data to be represented in different ways: Lucene for data discovery, Graph to find connections between data items, SQL for complex querying, NoSQL for retrieval from very large datasets, In-memory for ad-hoc data cubes etc etc. So why not avoid the question of which database is best and design your storage layer to represent the same data many different ways? We are in a data age, where the extraction of insight from data will be key to a company’s success. Storing data many ways will enable the greatest business benefit to be extracted and give companies which do a competitive edge. This presentation will discuss the merits and drawbacks of a greedy approach to data access and why this transforms the insights that can be extracted”.