90% of ML projects never make it to production. That's not a talent problem — it's an MLOps problem.

Phillip Mortimer is a computer scientist, ex-Chief Scientist and CTO at a London fintech, and one of the only people teaching MLOps at university level (Dauphine University, Paris — 5 years running).

In this episode, Phillip walks through the complete MLOps lifecycle:

• Data preparation — why EDA is the most forgotten step, and why data pipelines still matter in the LLM era

• Model building — Karpathy's 5-stage training cookbook: become one with your data → fit a baseline → overfit → regularise → squeeze out the juice

• Experiment tracking — MLflow, Weights & Biases, model registries, and model cards

• Deployment — real-time vs batch, Docker containers, inference optimisation with ONNX, vLLM, and TensorRT

• Monitoring — data drift, feedback loops, and keeping models relevant

• The future — why MLOps is shifting to AI engineering, and why agentic AI is the real breakthrough

Key stat: 90% of ML infrastructure cost is inference, not training. If you're not optimising your serving layer, you're burning money every day.

Your model only matters if it connects to the business. But when you're a data scientist learning MLOps on the fly, experimenting on live infrastructure is terrifying.

Anastasiia Kulakova is an Amsterdam-based data scientist at JetLakes, a fast-growing energy and mobility startup. She shares her candid journey from Jupyter Notebook to production-ready ML.

In this episode:

• Why "we have the data" from stakeholders rarely means what you think

• How to build your own MLOps learning sandbox without breaking production (GitHub Actions, Heroku, DigitalOcean)

• The reality of being a data generalist at a startup — wearing every hat from model training to Scrum Master

• How JetLakes uses predictive algorithms to balance the energy grid through EV charging optimisation

• Virtual power plants: turning parked electric vehicles into grid-scale flexibility

• Why the precision economy is coming to energy — and what that means for data teams

Just like a masterful orchestra needs a skilled conductor, successful data teams require leaders who can harmonize diverse talents into powerful performance. In this episode, top data leaders share their strategies for orchestrating success in modern data organizations.

0:00 - Introduction: The Orchestra of Data Leadership

1:27 - Meet Our Conductor-Level Leaders

2:13 - Bridging Academia & Industry with Ali Saad

4:33 - Practical Skill Development with Anastasia Kulakova

8:19 - Leadership Journey Insights with Julie Montel

11:15 - Data Storytelling Mastery with Promit Ray

13:49 - Key Takeaways & Action Steps

You'll discover:

• How to bridge the gap between academic knowledge and real-world application
• Practical strategies for gaining hands-on MLOps experience
• Multiple paths to data leadership positions
• Essential data storytelling techniques for stakeholder communication

Featured Data Leaders:

Ali Saad - Data Leader

Anastasia Kulakova - MLOps Expert

Julie Montel - Data Team Leader

Promit Ray - Data Science Leader

Ready to conduct your own data success story? Join our Data Chiefs community where we provide structured support, peer learning, and proven frameworks to help you advance your data leadership career.

Visit www.data-chiefs.com to start your journey.

#DataLeadership #DataScience #DataStrategy #CareerGrowth #MLOps

In this episode of Data Team Success, host Ross Webb sits down with Miltiadis Sarakinos, Head of Data Analytics at Bank Cler, to explore the fundamental strategies that create successful data leaders and thriving data careers. With years of experience in both scientific and business environments, Miltiadis shares invaluable insights on integrating data culture, overcoming data literacy challenges, and achieving career success in the ever-evolving field of data analytics.

Guest Introduction

Miltiadis Sarakinos brings a wealth of experience from his time at CERN and his current role as Head of Data Analytics at Bank Cler. His unique perspective bridges the gap between scientific rigor and practical business applications, making him an ideal guide for aspiring data professionals and leaders.

Key Takeaways

1. Establishing a Strong Data Culture

• The transition from scientific environments to business settings presents unique challenges for data professionals.
• Creating a data culture requires spreading knowledge and skills throughout the organization, not just within the data team.
• Every company is becoming a data and AI company, necessitating a broad understanding of data across all departments.

2. Overcoming Data Literacy Challenges

• Avoid creating data products that only the producers understand; focus on making insights accessible and actionable for the entire organization.
• Resist becoming solely a data delivery team; encourage other departments to develop their own data skills.
• Approach data management as an ongoing journey rather than a fixed project with a defined endpoint.

3. Building a Successful Data Career

• Master the fundamentals: Learn Python, SQL, and other in-demand skills by studying job postings and industry trends.
• Gain hands-on experience with real datasets, working through challenges and bugs on your own computer.
• Develop soft skills: Data roles are "50% marketing jobs," requiring the ability to communicate complex ideas to non-technical stakeholders.

4. Effective Team Management

• Balance individual creativity with standardization to ensure consistency and knowledge transfer within the team.
• Cultivate a lifelong learning mindset, continuously updating your skills and knowledge.

5. Overcoming Common Challenges

• Adapt to business needs by focusing on "80/20" solutions rather than striving for perfection.
• Maintain focus on key projects instead of getting distracted by multiple interesting but incomplete initiatives.
• Embrace iteration and continuous improvement in your work.

Final Advice: Become Data-Driven in Your Personal Development

Miltiadis emphasizes the importance of applying data-driven principles to your own career growth:

• Collect feedback and data about your performance and skills.
• Analyze this information objectively, without becoming defensive.
• Use these insights to continuously improve and grow as a data professional.

By following these strategies and maintaining a growth mindset, data professionals can build successful careers that adapt to the rapidly changing landscape of data and AI technologies.

In this episode of Data Team Success, we sit down with Annarita Roscino, Global Head of Data and Insight at Zurich Insurance Company. With over 20 years of experience in data leadership, Annarita shares invaluable insights on building and developing successful data teams, bridging technical and business perspectives, and navigating the complexities of ethical decision-making in data projects.

Guest Introduction

Annarita Roscino brings a wealth of experience to the table, having set up data teams from scratch throughout her career. Her journey from academia to industry leadership provides a unique perspective on the evolution of data roles and the challenges of building effective teams in the corporate world.

Key Insights

Building and Developing Data Teams

1. Diversity is Key: Annarita emphasizes the importance of building diverse teams, bringing together people from various backgrounds:

• Academia (PhDs)
• Business professionals with a passion for data
• Young talents from apprenticeship schemes
• Individuals with unconventional backgrounds (e.g., philosophy graduates with a knack for data)

1. Balancing Skills:

• Technical expertise is crucial, but commercial skills are equally important.
• The team needs to work at a business pace, not an academic one.

1. Evolving Team Structure:

• Started with data analysts, then data scientists.
• Recognized the need for data engineers, data quality experts, and data architects.

1. Hiring Beyond the CV:

• Look for passion and practical application of skills.
• Example: Hiring a philosophy graduate who built football prediction models in their spare time.

Bridging Technical and Business Perspectives

1. Alignment with Business Goals:

• Ensure data strategies are driven by business needs and pain points.
• Start with understanding the business strategy, then develop a data strategy to support it.

1. Collaboration is Crucial:

• Involve business stakeholders throughout the model-building process.
• Create transparency and build trust in data-driven decision making.

1. Quantifiable Impact:

• Tie data projects to tangible business outcomes.
• Example: Recovering costs from catastrophe claims to fund the entire data project.

1. Cross-functional Communication:

• Create frameworks that facilitate conversation between data experts and business professionals.
• Focus on creating a shared language and understanding.

Ethical Decision Making in Data Projects

1. Beyond Regulations:

• Ethics in data is about what you should do, not just what you can do legally.
• Example: The UK's GCSE grading model during COVID-19.

1. Questioning the Use of Data:

• Just because you can build a model doesn't mean you should.
• Consider the broader implications and potential biases.

1. Transparency and Explainability:

• Ensure models are transparent and understandable to those using them.
• Be aware of potential biases and work to mitigate them.

1. Ethical Framework:

• Develop a framework to assess and highlight potential ethical issues before deployment.
• Regularly review and update ethical guidelines as technology and society evolve.

Key Takeaways for Data Management Leaders

1. Cultivate Diversity: Build teams with varied backgrounds and skillsets to drive innovation and problem-solving.
2. Bridge the Gap: Create processes that facilitate ongoing communication between data teams and business units.
3. Focus on Impact: Align data strategies with clear business outcomes and quantify the value of data projects.
4. Ethical Consideration: Develop robust ethical frameworks and continuously question the implications of data use.
5. Adaptability: Stay flexible in your approach to team building and project management as the field of data analytics continues to evolve.

By implementing these strategies, data management leaders can build more effective teams, deliver greater value to their organizations, and navigate the complex ethical landscape of modern data analytics.

In this insightful episode, we're joined by Rimta Das, a senior data analyst at West Wing with a wealth of experience in navigating the complex world of data-driven roles. Rimta shares her journey from an introverted data provider to a confident data consultant, offering valuable strategies for aspiring data leaders looking to excel in their careers.

Key Takeaways for Data Management Leadership:

1. Mastering Stakeholder Communication:

• Present data insights in layman's terms, focusing on the situation, task, analysis, and result (STAR approach).
• Engage in open discussions with stakeholders to understand their needs and explore alternative solutions.
• Build relationships with stakeholders through one-on-one conversations to foster trust and collaboration.

1. Navigating Challenges with Senior Stakeholders:

• Advocate for data-driven approaches while acknowledging stakeholders' experience and business knowledge.
• Propose pilot projects or AB testing to demonstrate the value of data-driven initiatives without significant risk.
• Showcase the potential benefits of data-driven decision-making to gain buy-in from senior stakeholders.

1. Setting Realistic Expectations:

• Consolidate stakeholder requests into a single document and prioritize based on business impact and feasibility.
• Be transparent about project timelines and resource constraints to manage stakeholder expectations effectively.
• Develop a roadmap that aligns with business priorities and communicate it clearly to all stakeholders.

1. Anticipating Stakeholder Needs:

• Think proactively about additional questions or insights that may arise during presentations.
• Conduct thorough analyses and anticipate potential follow-up questions to avoid multiple iterations.
• Demonstrate a deep understanding of the business problem and provide comprehensive solutions.

1. Transitioning to a New Role or Domain:

• Prioritize stakeholder engagement and communication to understand their pain points and critical business needs.
• Take the initiative to build relationships, showcase your expertise, and demonstrate the value you bring to the organization.
• Continuously learn about the new domain and adapt your data-driven approaches to drive meaningful improvements.

By implementing these strategies, data management leaders can effectively navigate stakeholder relationships, set realistic expectations, and drive impactful data-driven initiatives. Rimta's insights serve as a valuable roadmap for professionals looking to advance their careers and become trusted data consultants within their organizations.

Don't miss this opportunity to learn from Rimta Das's extensive experience and elevate your data management skills. Tune in now and discover how you can unlock your full potential in the world of data-driven leadership!

In this insightful episode, we sit down with Rishabh Gupta, the Head of Group Data Science at Radius, to explore the evolving landscape of data management and leadership. With a wealth of experience in guiding organizations through their data journeys, Rishabh shares his expertise on tackling challenges, driving innovation, and paving the way for a successful career in data.

Key Takeaways for Data Management Leaders:

Shifting Focus in Data Management:

Organizations are realizing the importance of robust data infrastructure as a foundation for advanced analytics and AI applications.

Investing in data quality, data governance, and modern data tools is crucial for unlocking the true potential of data.

Overcoming Challenges and Securing Buy-in:

Demonstrating the value of data initiatives through small pilots and proof-of-concepts can help gain stakeholder support.

Aligning data projects with business pain points and showcasing their impact on revenue, efficiency, or insights can accelerate buy-in.

Mastering Data Classification:

Data classification is a critical but often challenging task, especially in complex organizational structures with siloed data.

Collaborating with cross-functional teams, mapping data flows, and prioritizing high-value datasets are key strategies for effective data classification.

Driving Adoption and Understanding:

Effective communication and change management are vital for fostering a data-driven culture.

Breaking down complex technical concepts into simple terms and highlighting the benefits for each stakeholder can drive adoption and understanding of data initiatives.

Embracing Emerging Technologies:

Data Fabric and similar technologies show promise in addressing data silos and enabling seamless integration across disparate data sources.

Practical applications of AI that solve real-world problems and enhance day-to-day operations will be the true game-changers.

Proactive Career Growth in Data:

Continuously learning, staying updated with the latest trends, and applying data science techniques to real-world scenarios are essential for career growth.

Developing strong communication skills, storytelling abilities, and the art of presenting insights to non-technical audiences will set you apart as a data leader.

By implementing these strategies and embracing a proactive approach to data management and leadership, you can navigate the complexities of the data landscape, drive innovation within your organization, and accelerate your career growth in this exciting field.

In this episode, we dive into the world of data governance with Uwe Scheid, a seasoned data and analytics director at Mindshare. Uwe shares his wealth of experience and practical strategies to help data management leaders navigate the complexities of digital transformation and drive their careers forward.

Key Takeaways:

1. Effective Communication: Simplify complex data topics and explain them in a way that everyone can understand. Focus on the benefits of change and how it will positively impact workflows and processes.

2. Workflow Changes: When implementing new data processes, clearly communicate the reasons behind the changes and how they will benefit the organization. Gain support from management to ensure a smoother transition.

3. Data Governance: Start small and gradually expand your data governance initiatives. Continuously adapt to the ever-changing digital landscape and treat data governance as an ongoing process rather than a one-time project.

4. Integrating AI: Embrace the pull of AI technology, especially among younger workforce members. Provide a safe environment for employees to work with AI and democratize access to these tools.

5. Continuous Learning: Stay ahead in the rapidly evolving world of data and analytics by constantly learning and seeking inspiration from various industries. Develop a broad understanding of tools, systems, and algorithms to thrive in this dynamic field.

By applying these actionable insights, data management leaders can effectively navigate the challenges of digital transformation, drive change within their organizations, and propel their careers to new heights. Embrace the power of effective communication, adaptable data governance strategies, and a commitment to continuous learning to unlock your full potential in the world of data and analytics.