Value Driven Data Science

The success of data science projects often depends on being able to get stakeholders, from a variety of backgrounds, to work well together. But what if the stakeholders involved come from very different backgrounds and struggle to understand each other – as can be the case with data scientists and engineers? 

In this episode, Dr Genevieve Hayes is joined by software engineer turned data scientist Hendrik Dreyer, who has carved a niche for himself by acting as a intermediary between Team Data Science and Team Engineering.

Guest Bio

Hendrik Dreyer is both a qualified data scientist and a qualified engineer. He worked extensively in a range of senior software engineering roles, in both South Africa and Australia, prior to making the transition into data science. He is now the Manager of Analytics Capability at Australia’s largest superannuation fund, AustralianSuper.

Talking Points

• The different mindsets commonly held by data scientists, data engineers and the business in general, when it comes to data science and analytics.
• How these diverse mindsets can give rise to challenges, when it comes to delivering data science solutions, and the potential consequences if these challenges aren’t adequately addressed.
• Approaches to bridging the chasm between data science, data engineering and the business.
• Actions each of these different groups of people can take in order to help bridge the divide within their own organisations.
• The unexpected benefit of agile project management as a people development tool.

Links

• Connect with Hendrik on LinkedIn
• Connect with Genevieve on LinkedIn
• Be among the first to hear about the release of each new podcast episode by signing up HERE

Businesses rarely approach data scientists with well-defined problems to solve. Sometimes, the problems businesses devise aren’t appropriate for solving using data science at all. This makes it difficult for data science projects to succeed. 

In this episode, Dr Genevieve Hayes is joined by Rob Deutsch to discuss strategies businesses and data scientists can employ to identify data science use cases and maximise their probability of success.

Guest Bio

Rob Deutsch is the Chief Operating Officer of AkuShaper, a company that uses advanced modelling algorithms and software to build better surfboards faster. He is also a data science consultant with Parity Analytic, and previously founded Boxer, which built software for creating better financial models.

Talking Points

• Processes for identifying and understanding business problems, and determining whether a data science solution is appropriate and what that solution should look like.
• The different ways in which people from different backgrounds can look at a data science problem and how that influences the questions they ask of data/the way they tackle problems.
• The role of the business vs the role of the data scientist in defining/scoping data science projects.
• How to maximise the probability of success of a data science project.
• How the data science/data analytics skill set can be transferred to areas outside of technical data analysis, such as running a SaaS company.

Links

• Connect with Rob on LinkedIn
• Bird App XKCD Comic
• Connect with Genevieve on LinkedIn
• Be among the first to hear about the release of each new podcast episode by signing up HERE

Have you ever wondered what your organisation’s Board are thinking, when it comes to data use?

In this episode, Dr Genevieve Hayes is joined by Dr Stuart Black to discuss the attitudes of Boards to data use and their implications for the organisations they govern.

Guest Bio

Dr Stuart Black is an Enterprise Fellow in data, analytics, disruption and innovation at the University of Melbourne. Prior to joining academia, Stuart spent 30 years in professional services and industry, at employers including Deloitte, where he was Senior Partner, National Australia Bank and AT Kearney. He is also a co-author of the recently released book Business Model Transformation – the AI and Cloud Technology Revolution.

Talking Points

• The Board’s role in catalysing and controlling data-driven business model transformation.
• What is meant by the secondary use of data and what are some of the opportunities and threats presented by it?
• Why intellectual curiosity is more important than prior data experience in maximising the competitive advantage of data.
• The importance of taking a medium term view when it comes to data initiatives.
• The key Board attributes that determine an organisation’s attitudes towards data as an enabler of strategy.
• Strategies for shifting the attitude of your Board in order to encourage a mindset that is more supportive of data initiatives.

Links

• Connect with Stuart on LinkedIn
• Stuart’s University of Melbourne Profile
• Business Model Transformation – the AI and Cloud Technology Revolution Microsite
• Connect with Genevieve on LinkedIn
• Be among the first to hear about the release of each new podcast episode by signing up HERE

We all know what it means for a human to discriminate against another human, but the concept of a predictive model or an artificial intelligence is relatively new. What does it mean for a model or an AI to discriminate against someone? 

In this episode of Value Driven Data Science, Dr Genevieve Hayes is joined by Dr Fei Huang to discuss the importance of considering fairness and avoiding discrimination when developing machine learning models for your business.

Guest Bio

Dr Fei Huang is a senior lecturer in the School of Risk and Actuarial Studies at the University of New South Wales, who has won awards for both her teaching and her research. Her main research interest is predictive modelling and data analytics, and has recently been focussing on insurance discrimination and pricing fairness.

Talking Points

• Direct vs indirect discrimination and how data scientists can create discriminatory machine learning models without ever intending to.
• What it means for a model to be fair and the trade-off that exists between individual and group fairness.
• How fairness and discrimination come up (and have been addressed) in different applications of machine learning, including (but not limited to) insurance.
• How different jurisdictions are currently addressing algorithmic discrimination, through regulation and other means.
• What this means for organisations who currently make use of machine learning models or would like to in the future.
• Why organisations should start considering fairness and discrimination when using analytics and what they can do about it now.

Links

• Connect with Fei on LinkedIn
• Fei’s UNSW Public Profile
• Australia’s Voluntary Ethical AI Framework

Fei’s papers on fairML and insurance pricing:

• Anti-Discrimination Insurance Pricing: Regulations, Fairness Criteria and Modelling
• The Discriminating (Pricing) Actuary
• Welfare Implications of Fairness and Accountability for Insurance Pricing
• Connect with Genevieve on LinkedIn
• Be among the first to hear about the release of each new podcast episode by signing up HERE

The two most challenging transitions you can make in your career are transitioning from individual contributor to team lead, and moving from team lead to managing managers. This is true across all professions, but is particularly pronounced in technical fields, like data science. 

In this episode, host Dr Genevieve Hayes is joined by guest Tim Davey to discuss the challenges faced by data scientists looking to climb the corporate ladder, and how employers of data professionals can support them in developing their careers.

Guest Bio

Tim Davey has spent the majority of his career working in the organisational development and HR space where his work has focussed strongly on the development of leaders and working with individuals to understand and maximise their careers. This has included, among other things, providing executive coaching to senior management across a wide range of industries, including media, the performing arts, manufacturing, financial services, transport, education, insurance, legal, and not-for-profit sectors. 

Yet, Tim also has a strong technical background himself, having completed a Science degree at the University of Melbourne, and starting his working career in the chemical manufacturing sector, so has first-hand understanding of the challenges faced by the members and leaders of technical teams.

Talking Points

• The key differences between working as an individual contributor vs line manager vs senior manager.
• Why people can struggle to make the transition between data scientist and team lead and what can be done to make it easier.
• The importance of technical capability vs managerial skills in technical leadership roles, and how organisations can support staff to develop those skill sets if one is lacking or weaker.
• Managing a team and building your profile in the post-COVID, remote working world.
• Advice for data scientists considering moving into managerial roles – and for those who would prefer to remain an individual contributor.

Links

• Connect with Tim on LinkedIn
• Connect with Genevieve on LinkedIn
• Download the FREE Data Science Project Discovery Guide
• Genevieve Hayes Consulting offers one-on-one coaching for new and aspiring data science and analytics leaders. To find out more, or to share your thoughts and feedback on the podcast, you can get in touch here.
• Be among the first to hear about the release of each new podcast episode by signing up HERE

Data presents incredible opportunities for organisations to create value, but with the current skills and labour shortages that are affecting all businesses, finding and retaining data scientists and other data professionals can be hard. 

In this episode, host Dr Genevieve Hayes is joined by guest Amanda Aitken to discuss a practical way in which organisations can address the skills shortage, gain much needed data skills and increase staff retention – by upskilling their existing staff.

Guest Bio

Amanda Aitken is a fully-qualified actuary who is currently an educator with the Actuaries Institute of Australia. She teaches data analytics and data science to actuaries through the Actuaries Institute’s Data Analytics Application course and is also a member of the Institute’s Data Analytics Practice Committee and Data Analytics Education Faculty.

Talking Points

• The difference between an actuary and a data scientist.
• Why data skills are becoming increasingly important and the benefits to organisations of upskilling their data team.
• What’s involved in upskilling data-focussed staff.
• The importance of considering “soft skills”, such as communication, privacy and ethics, when training data scientists.
• How an organisation can get the most out of their data staff once they have been upskilled.

Links

• Connect with Amanda on LinkedIn
• Data Science Applications Microcredential
• Data Analytics Seminar
• Machines Behaving Badly by Toby Walsh
• How Humans Judge Machines by Cesar Hidalgo
• Download the FREE Data Science Project Discovery Guide
• Connect with Genevieve on LinkedIn
• To find out more about custom data science training from Genevieve Hayes Consulting or to share your thoughts and feedback on the podcast, you can get in touch HERE.
• Be among the first to hear about the release of each new podcast episode by signing up HERE