BossNData

In this episode of the Boss N' Data Podcast, Joshua Smith is joined by Andrea Rosales, PhD, Associate Director of Data Science at Blend.

Andrea has spent more than fifteen years across academia and industry. Her PhD at the University of St Andrews looked at how AI systems learn and adapt in messy, real-world conditions, and her work has moved from transfer and continual learning through to the reasoning and reliability of large language models. She is also co-founder of Mosaic, a Scotland-based community focused on AI governance, responsible AI and inclusive tech.

This is a conversation about the uncomfortable middle ground: not "AI will take your job" or "AI changes nothing", but what actually shifts when the modelling gets automated, the org chart changes shape, and nobody can quite say who signed off on the model.

You'll hear:

• The difference between robustness and reliability
• Why the data science bottleneck has moved from modelling to deployment, and what that does to how you build a team
• Andrea’s thoughts on what happens to junior roles when the repetitive work is automated
• Andrea's concern about critical thinking, from students who can't explain their own essays to developers who no longer own the reasoning behind their code
• What we're outsourcing besides intelligence
• Agent washing explained
• The agentic productivity paradox and how it plays out
• Three practical tests for meaningful AI accountability
• Data for good and the human-centric metrics that prove it

Disclaimer: The views expressed by the guest in this episode are her own and do not reflect those of her current or former employers.

Andrea Rosales, PhD, is Associate Director of Data Science at Blend, and an AI research and innovation lead with over fifteen years' experience delivering data and AI solutions across industry and academia.

Andrea holds a PhD in computer science from the University of St Andrews, where her research focused on how AI systems learn and adapt in complex real-world environments. Her work has evolved from domain adaptation and continual learning into research on large language model reasoning and reliability, with a strong emphasis on building AI systems that are explainable and robust.

She now works with organisations to design and deliver end-to-end AI solutions, with a focus on responsible AI adoption and data for good. She is co-founder of MosAIc, a Scotland-based community for people working on AI governance, responsible AI and inclusive tech.

👉 Connect with Andrea on LinkedIn: https://www.linkedin.com/in/andrea-rosales-phd

👉 Find out more about MosAIc Scotland: https://mosaic-community.uk/

Joshua Smith has been a Data and AI talent specialist since 2011. He is Sales Director at MBN Solutions, responsible for leading and supporting the growth and diversification of MBN's client portfolio, and building MBN's position as a leading Data & AI talent consultancy across the UK, Europe, and US.

Explore industry updates and the latest in data and AI talent on MBN Solutions' LinkedIn page.

Hosted on Acast. See acast.com/privacy for more information.

In this episode of the Boss N’ Data Podcast, Joshua Smith sits down with the MetaMetrics team ,Rebecca Shaw (Client Services Manager), Sam Watts (Director) and Philip Gaudoin (Director) for an honest conversation about the modern measurement paradox: marketers have more data than ever, and confidence in decisions has rarely been lower.

MetaMetrics is a specialist consultancy that has been doing econometric modelling (also known as marketing mix modelling) for nearly twenty years. They’ve seen the apparent death and the very real rebirth of econometrics, and they have strong views on why it has outlasted every shiny new thing thrown at it.

You’ll hear:

• Why the data problem changed but the underlying stats really hasn’t
• Analysis paralysis: when more data made marketers less confident, not more
• Measurement vs application: when the problem isn’t the model, it’s the adoption
• The honest take on where AI actually helps in marketing measurement and where it doesn’t
• Phil’s warning to stop outsourcing your thinking
• Vanity metrics, PPC and the new role of LLM impression data
• Brand vs performance: why econometrics is no silver bullet for brand and the client who lost 30p in every £1 to negative PR
• How to build real confidence with senior stakeholders

Disclaimer: The views expressed by the guests in this episode are their own and do not reflect those of their current or former employers.

Philip Gaudoin: Director, MetaMetrics

Phil joined MetaMetrics in 2016, bringing a commercial, consulting focus to the work. He started his career as an academic at the University of Cambridge teaching econometrics to undergraduates, before moving into marketing analytics in 2005 at Publicis Groupe. He’s been pulling threads between maths, marketing and management ever since.

👉 Connect with Phil on LinkedIn: https://uk.linkedin.com/in/philip-gaudoin-44714b44

Sam Watts: Director, MetaMetrics

Sam has worked in econometrics his entire career — first across media agencies, including Publicis Groupe, before joining MetaMetrics around 15 years ago. He leads delivery at MetaMetrics, with particular depth in the evaluation of digital and social channels and the pragmatic use of attribution modelling for clients across multiple markets.

👉 Connect with Sam on LinkedIn: https://www.linkedin.com/in/samuelawatts/

Rebecca Shaw: Client Services Manager, MetaMetrics

Rebecca came to econometrics from a marketing background, with a long-standing interest in marketing effectiveness and what truly drives performance. She supports MetaMetrics clients through the full lifecycle — making sure the analysis is digestible, the insights translate to action, and the relationship outlasts the project.

👉 Connect with Rebecca on LinkedIn: https://www.linkedin.com/in/rebecca-shaw-3752a711b/

Joshua Smith has been a Data and AI talent specialist since 2011. He is Sales Director at MBN Solutions, responsible for leading and supporting the growth and diversification of MBN’s client portfolio, and building MBN’s position as a leading Data & AI talent consultancy across the UK, Europe, and US.

Explore industry updates and the latest in data and AI talent on MBN Solutions’ LinkedIn page.

Hosted on Acast. See acast.com/privacy for more information.

In this episode of the Boss N’ Data Podcast, Joshua Smith is joined by Sean Harkin, PhD, Data Science & AI Lead at Lloyds Banking Group, for conversation about what it actually takes to move AI from proof-of-concept to production in a highly regulated environment.

Sean has 18 years across financial services from risk modelling in investment banks, through fintech, to leading data science and AI in the insurance arm of Lloyds Banking Group. He is currently shipping GenAI and agentic solutions into production where most firms get stuck.

You’ll hear:

• How Sean’s team controls LLM hallucination at production quality ontologies
• Why decision-recommendation tools in underwriting are designed to support humans, not replace them
• When to reach for a domain-specific model versus careful prompt engineering on a general LLM
• How to make AI reasoning transparent enough for regulators, risk teams, and reinsurers
• The cost-per-token reality check on agentic systems most teams underestimate
• What separates AI projects that ship from POCs that stall and the customer journey result one of Sean’s use cases demonstrates
• Why doing the science well and doing risk management well are the same thing

Disclaimer: The views expressed by the guest in this episode are their own and do not reflect those of their current or former employers.

Sean Harkin, PhD, is Data Science & AI Lead at Lloyds Banking Group, working in the insurance arm. He has 18 years’ experience in financial services across startups, banks, and consultancies, with a research science background spanning biochemistry, biophysics, and a PhD in applied statistical modelling focused on causal inference.

Sean now leads a team of data scientists, ML engineers, and other technical professionals putting GenAI and agentic solutions into production, cutting through the POC bottleneck that holds many firms back. His current work focuses on AI for medical documentation analysis, decision recommendation, and the practical realities of shipping AI safely in a regulated environment.

👉 Connect with Sean on LinkedIn: https://www.linkedin.com/in/sean-harkin-phd-22776674/

Joshua Smith has been a Data and AI talent specialist since 2011. He is Sales Director at MBN Solutions, responsible for leading and supporting the growth and diversification of MBN’s client portfolio, and building MBN’s position as a leading Data & AI talent consultancy across the UK, Europe, and US.

Explore industry updates and the latest in data and AI talent on MBN Solutions’ LinkedIn page.

Hosted on Acast. See acast.com/privacy for more information.

In this episode of the Boss N’ Data Podcast, Joshua Smith is joined by Christopher Conroy, Chief Data & Product Officer at Future Anthem, for a proper behind-the-scenes look at what it takes to scale a data function and a platform at the same time.

Chris has lived the full journey: from the early analytics → data science wave (2014/2015) into building Future Anthem over the last six and a half years. 

• What pulled Chris into building real-time personalisation in betting/gaming
• How the early analytics → data science wave (2014/15) compares to what’s happening now
• How the technical function at Future Anthem evolved from day one to today
• How they thought about generalists vs specialists early on 
• What roles became essential once they hit a certain scale
• What architectural inflection points showed up as the platform scaled and what needed change

Disclaimer: The views expressed by the guest in this episode are their own and do not reflect those of their current or former employers.

About the Guest

Christopher Conroy is the Chief Data & Product Officer at Future Anthem, a real-time personalisation platform in the betting and gaming industry. He’s spent the last six and a half years building the technical function and embedding product at Future Anthem, focused on real-time AI systems designed to operate at scale, with performance targets measured in milliseconds. 

Before Future Anthem, Christopher worked across the early wave of analytics → data science transformation, including roles at Rank Group, where cloud and machine learning capabilities were being built out at enterprise level.

 👉 Connect with Chris Conroy on LinkedIn: https://www.linkedin.com/in/chrisconroydata/

About Your Host

Joshua Smith has been a Data and AI talent specialist since 2011. He is Sales Director at MBN Solutions, responsible for leading and supporting the growth and diversification of MBN’s client portfolio, and building MBN’s position as a leading Data & AI talent consultancy across the UK, Europe, and US. 

Explore industry updates and the latest in data and AI talent on MBN Solutions’ LinkedIn page.

Hosted on Acast. See acast.com/privacy for more information.

In this episode of the Boss N’ Data Podcast, Joshua Smith is joined by Rich Hunt, Director of Data & Insight at Openreach, to unpack what it really takes to lead data teams through constant change. 

Openreach’s team has won DataIQ Best Data & AI Team of 2025 and in this episode, Rich shares how Openreach approaches transformation, from moving to a hub-and-spoke analytics model, to navigating a major migration and keeping teams focused on outcomes.

They dig into the realities data leaders face: long timelines, rising costs, stakeholder skepticism, and the risk of data teams becoming detached from business value. 

They also discuss GenAI skepticism vs hype, and why building change resilience is now part of the job.

By the end of this episode, you’ll understand:

• What Openreach does and why its data challenges are uniquely complex
• How to structure data teams to stay close to decision-making while scaling capability
• How to defend long-term data investment when ROI isn’t instant
• How to keep data work relevant: translating business priorities into data-led impact
• Agentic AI: cutting through hype and skepticism
• What change resilience looks like inside a complex, regulated organisation.
• Practical advice for leading teams through change
• The next big shift for data leaders to prepare teams for

Disclaimer: The views expressed by the guest in this episode are their own and do not reflect those of their current or former employers.

About the Guest

Rich Hunt is the Director of Data & Insight at Openreach, a senior data leader with 16+ years’ experience. He leads end-to-end data capability and has been accountable for connecting business strategy to data-led solutions, including leading Openreach’s migration to cloud and supporting cultural change around data. Openreach’s team has won DataIQ Best Data & AI Team of 2025.

 👉 Connect with Rich Hunt on LinkedIn: https://www.linkedin.com/in/huntr/

About Your Host

Joshua Smith has been a Data and AI talent specialist since 2011. He is Sales Director at MBN Solutions, responsible for leading and supporting the growth and diversification of MBN’s client portfolio, and building MBN’s position as a leading Data & AI talent consultancy across the UK, Europe, and US. 

Explore industry updates and the latest in data and AI talent on MBN Solutions’ LinkedIn page.

Hosted on Acast. See acast.com/privacy for more information.

In this episode of the Boss N Data Podcast, host Joshua Smith sits down with Chris Mitchell, Head of Data & Marketing Operations at Sky Business and former WPP agency leader, to unpack how world-class organisations are building marketing operations that actually drive growth.

With over two decades bridging data, marketing, and operations, Chris has helped major brands embed data-driven culture, reduce churn, and align analytics, BI, and marketing into one cohesive growth engine.

From offline segmentation and direct mail to AI and automation, his career charts the full evolution of how businesses use data to make smarter, faster decisions.

By the end of this episode, you’ll understand:

• Why most companies don’t have a data problem - they have a usage problem
• How Sky uses marketing operations as a growth engine by uniting BI, data science, and execution
• The simple model that helped Sky reduce churn through data-led customer engagement
• What to look for (and avoid) when investing in MarTech and AI tools
• Why AI needs people and how to use it to enhance, not replace, decision-making

About the Guest

Chris Mitchell is an innovative and enthusiastic leader with over two decades of experience delivering data and digital transformation across multiple sectors.

Chris advocates for using data as a strategic growth engine - driving incremental revenue, customer engagement, and organisational adoption of best-practice frameworks.

His approach has underpinned award-winning marketing campaigns and helped major brands embed a culture where data is at the heart of every decision..

👉 Connect with Chris on LinkedIn: https://www.linkedin.com/in/chris-mitchell-49636a4/

About Your Host

Joshua Smith has been a Data and AI talent specialist since 2011. He is Sales Director at MBN Solutions, responsible for leading and supporting the growth and diversification of MBN’s client portfolio, and building MBN’s position as a leading Data & AI talent consultancy across the UK, Europe, and US. 

Explore industry updates and the latest in data and AI talent on MBN Solutions’ LinkedIn page.

Disclaimer

The views expressed by the guest in this episode are their own and do not reflect those of their current or former employers.

Hosted on Acast. See acast.com/privacy for more information.

In this episode of the Boss N Data Podcast, host Joshua Smith sits down with Harry Davies, VP of Marketing Investment & Effectiveness at Sage, formerly at Google and Amazon, to unpack the truths about marketing effectiveness.

Harry shares invaluable insights from a career spanning government behaviour change campaigns, launching Amazon Marketing Services in Europe, and leading measurement and data science at Google Cloud. Today at Sage, he’s helping drive £2.1Bn in recurring revenue by shifting from pure demand gen to a balanced brand + activation approach.

By the end of this episode, you’ll understand:

• Why sales and marketing so often pull in different directions - and the practical fix.
• How to measure what really works in marketing (beyond vanity metrics).
• The power of combining experiments, econometrics, and attribution.
• How buyer behaviour is changing and what it means for sales and marketing leaders.
• Where AI is already making marketing effectiveness faster and smarter.
• Why you should start small, solve frontline challenges, and scale from there.

About the Guest

Harry Davies is an experienced Marketing Strategy and Operations leader, currently VP of Marketing Investment & Effectiveness at Sage. He previously led measurement and data science teams at Google and Amazon and sits on the Advisory Board of the International Journal of Market Research. Harry is a regular speaker on marketing strategy, science, and effectiveness.

👉 Connect with Harry on LinkedIn: https://www.linkedin.com/in/harrydavies/

About Your Host

Joshua Smith has been a Data and AI talent specialist since 2011. He is Sales Director at MBN Solutions, responsible for leading and supporting the growth and diversification of MBN’s client portfolio, and building MBN’s position as a leading Data & AI talent consultancy across the UK, Europe, and US. 

Explore industry updates and the latest in data and AI talent on MBN Solutions’ LinkedIn page.

Disclaimer

The views expressed by the guest in this episode are their own and do not reflect those of their current or former employers.

Hosted on Acast. See acast.com/privacy for more information.

Bias in AI isn’t just a technical glitch - it’s one of the hardest problems in machine learning today. While many organisations treat it as a regulatory checkbox, the reality is that unchecked bias can damage trust, weaken systems, and cost businesses far more than they realise.

In this episode, host Joshua sits down with Raluca Crisan - data scientist, co-founder, and CTO of Etiq, to explore the real challenges of bias in AI, and what it takes to build systems that are ethical, robust, and genuinely useful.

Recognised as Women in AI Europe 1st Prize (2020) and named among the 100 Women in AI Ethics (2021), Raluca brings a rare perspective on how bias, adoption, and ethics intersect in real-world AI. In this conversation, she shares practical lessons on tackling bias, the pitfalls of “big data” buzzwords, and what the rise of copilots and agents means for the future of hiring.

In this episode, you’ll learn:

• Why bias in AI is more than a “checkbox” issue and how to address it meaningfully.
• The biggest adoption problem most AI testing tools overlook.
• How to balance commercial pressure with purposeful, ethical outcomes.
• What recruiters and data leaders need to know about new AI roles.
• The hardest part of doing something new in AI today from a founder’s perspective.

If you found this episode useful, don’t keep it to yourself - share it with your network, subscribe on your favourite platform, and leave us a quick review so more data professionals can find the show.

Subscribe & Follow: Enjoying the show? Don’t miss an episode. Follow the podcast wherever you get your podcasts and leave us a review to help more listeners discover these conversations.

Connect with MBN Solutions:

MBN Solutions on LinkedIn – Explore industry updates, groundbreaking projects, and the latest in data and analytics talent. Join the conversation and discover how MBN Solutions is shaping the future of data leadership. Visit the company page here.

Visit MBN Solutions website at www.mbnsolutions.com

Hosted on Acast. See acast.com/privacy for more information.

In this episode I am joined by Dr Shorful Islam, co-founder and CEO of Be Data Solutions.

With over 20 years of experience in the data and AI industry, Dr. Islam has been applying machine learning since his Psychology PhD—before “data science” became mainstream. Since then, he has helped global companies develop data strategies and build predictive models that drive business growth.

He specialises in building data functions across organisations such as ITV and Wunderman and is the author of Data Culture, a book on integrating data as a fundamental organisational component.

Beyond his leadership at Be Data Solutions, Dr. Islam founded Subatomic Analytics and advises multiple startups on data and AI. His career includes roles as Chief Data Scientist at Tribal Worldwide, Managing Director at Stream Intelligence, and Managing Partner at Wunderman, along with leadership positions at adam&eveDDB/DDB and ITV. He also brings experience from the NHS and local government.

Hosted on Acast. See acast.com/privacy for more information.

Hosted on Acast. See acast.com/privacy for more information.