The Double Trio Framework for AI Product Management

Structuring Your AI Initiatives and Your Learning Path to Becoming an AI PM

💡 This framework will be steadily improved. You are now reading The Double Trio Framework v2.0 for AI Product Management. You can read about the changes here:

Changelog 📃 The Double Trio Framework v.2.0 for AI Product Management

---

#beyondAI

"It’s so complex I don’t even know where to start." - This sentiment is absolutely true!

Understanding where to start is a common struggle for aspiring AI Product Managers (AIPMs), and even professionals regularly face the challenge of determining what to tackle next. The fear of overlooking important aspects of their AI product initiatives is real and can be daunting.

And it’s obvious what might happen:

1. Aspiring AIPMs might give up too soon on their learning journey.

2. Professionals might develop a subpar AI product.

This was me as well, some years ago, until I decided to consolidate everything I had learned and develop the Double Trio Framework.

In this issue, you will learn:

• What the Double Trio is

• Why the Double Trio Matters

• Double Trio Deep Dive:

  • Technical Trio

  • Strategic Trio

• Balancing the Double Trio

Happy reading 🛋️

---

---

What is the Double Trio?

You might already be familiar with my hashtag beyondAI. It adorns every post I write on LinkedIn and headlines my Substack issues. It encapsulates my philosophy of AI Product Management. Building successful AI products goes beyond merely leveraging AI technology, it extends far beyond the technical aspects. It is a blend of various fields, methods, and approaches.

To simplify this complex process, I use a framework to guide my thinking from the start of any AI product ambition.

I call this framework the Double Trio.

The Double Trio consists of three elements focused on the technical aspects of an AI Product - Data, IT, and AI itself - and three elements that ensure these technologies are effectively implemented within a business context - Governance, Business, and People. This duality highlights the essential balance between technology and the human-centric approaches needed for successful AI products.

To build an AI Product, one must master not only the technological framework but also the operational and strategic frameworks that make AI solutions viable, ethical, and user-friendly. Each trio complements the other and creates a holistic model that addresses the full spectrum of challenges and opportunities in AI product development.

It always served me as a constant reminder that in AI Product Management, success stems from harmonizing these six elements.

Leaving out any of these elements means your AI product ambition will not survive, let alone thrive.

Each of these areas encompasses a vast range of topics, and mastering them might take years.

But you don't need to be an expert in all these areas; you just need to understand enough to recognize the need for expert support on this journey.

There is one thing you must master:

AI Product Management - the art and science of bringing together the right experts to build an AI solution for a real problem that people are willing to pay for.

That’s our expertise.

Why the Double Trio matters?

My aim in introducing this framework to you has various reasons. First, I think it might help you understand how complex building AI products are and what you need to consider once you start your AI initiative.

The six dimensions of the Double Trio—Technical Trio (IT, Data, AI) and Strategic Trio (Business, Governance, People)—together describe the comprehensive complexity of AI product development. Each dimension represents a critical area that must be addressed and balanced to achieve success.

Second, it helps to structure the way of working as an AI Product Manager.

For example, whether I'm creating a slide deck or planning a project, I ensure to cover all six elements, if necessary. It immediately structures my thoughts and helps me remember what we need to cover. I even name Jira tickets like this:

1. [Governance] - Incorporate Data Governance guidelines into the Training Data Pipeline

2. [Business] - Conduct Market Research for New AI Product Launch

3. [People] - Develop AI Product Training Program for Users

4. [IT] - migrate to new Cloud Infrastructure for Enhanced AI Processing

But it will also serve as the guiding structure for the AI Product Management Learning Repository I’m building. I believe a repository should be easy to navigate to support and serve its purpose. Even though we can easily search for content using the search functionality , having intuitive navigation is essential for those who don’t know exactly what to search for. An easy-to-navigate structure can support users in finding the information they need more efficiently.

Let’s delve into each dimension in more depth.

---

The Technical Trio: Data, IT, AI

The first trio—Data, IT, and AI—forms the backbone of your AI product's functionality. With these three elements, you can build a fully functioning AI solution.

Data

The quality, breadth, and depth of data determine the potential of what our AI can achieve. Data is the knowledge foundation of any AI solution. Understanding AI Product Management means recognizing that data encompasses the collection, storage, processing, and analysis of data. Equally important, but often overlooked, are data management and governance, which ensure data is handled effectively and responsibly.

These are some topics we need to consider when capturing the essence of Data:

• Data Collection: Gathering data from various sources, such as databases, APIs, and sensors.

• Data Storage: Ensuring data is stored securely and efficiently, using databases and cloud services.

• Data Processing: Cleaning and transforming raw data into a usable format.

• Data Quality: Ensuring data is accurate, complete, and reliable for AI model training.

• Data Integration: Combining data from different sources to create a comprehensive dataset.

• Data Governance: Establishing policies and procedures for managing data assets responsibly.

• Data Lifecycle Management: Overseeing the data from its creation and storage to its eventual archiving and deletion.

• Big Data Technologies: Utilizing tools and platforms to handle large volumes of data effectively.

IT

When I talk about IT in the context of AI Product Management, I refer to the software needed to make AI accessible, the infrastructure that interconnects data-producing, data-processing, and data-delivering systems, and the knowledge required to understand that every AI endeavor relies on a reliable IT ecosystem. This ecosystem must deliver scalable and secure AI solutions. Everything in AI builds on top of IT systems. Understanding these overall concepts helps plan your AI initiative more accurately.

These are some topics we need to consider when capturing the essence of IT:

• Software Development: Creating applications and tools that facilitate AI accessibility and usability.

• Infrastructure: Setting up servers, cloud services, and networks to support AI workloads.

• Data Interconnectivity: Ensuring seamless integration and communication between data-producing, data-processing, and data-delivering systems.

• Scalability: Designing IT systems that can grow and handle increased demand as AI projects expand.

• Security: Implementing robust security measures to protect data and AI systems from threats.

• Cloud Computing: Leveraging cloud platforms to provide flexible and scalable resources for AI development and deployment.

• IT Governance: Aligning IT strategy with business goals and managing IT-related risks effectively.

• Monitoring and Maintenance: Continuously overseeing IT systems to ensure optimal performance and quickly addressing any issues.

AI

This component is where the magic happens—the algorithms that drive decision-making and learning within the product. This is why AI in AI products is so prominent. It's because of this element that we call it an AI product and not just a software product. The main solution comes from the AI technology. Navigating through a diverse set of AI technologies, understanding which is suitable for which problem set, and recognizing their commonalities is absolutely essential. This knowledge enables you to communicate effectively with various stakeholders, make informed decisions, and uncover opportunities.

These are some topics we need to consider when capturing the essence of AI:

• Machine Learning: Algorithms that enable systems to learn from data and improve over time.

• Deep Learning: Advanced neural networks capable of handling complex tasks like image and speech recognition.

• Natural Language Processing (NLP): Techniques for understanding and generating human language.

• Computer Vision: Methods for interpreting and analyzing visual information.

• Predictive Analytics: Using historical data to make informed predictions about future events.

• Prescriptive Analytics: Recommending actions based on predictive analytics to achieve desired outcomes.

• Generative AI: Creating new content or data that resembles existing data, such as text, images, or music.

---

The Strategic Trio: Governance, Business, People

While the first trio—the Technical Trio—is absolutely essential to build the AI solution, it becomes irrelevant if you don’t pay equal attention to the Strategic Trio. Governance, Business, and People ensure that your AI solution can evolve into a successful AI product.

Governance

When I refer to governance, I mean the various types needed to ensure everything happens in a standardized, ethical, privacy-compliant, and secure manner. This includes all forms of governance, whether it be AI, data, or IT governance.

These are some topics we need to consider when capturing the essence of Governance:

• Regulatory Compliance: Adhering to laws and regulations relevant to AI and data use.

• Ethical Guidelines: Ensuring AI systems are fair, transparent, and respect user privacy.

• Risk Management: Identifying and mitigating potential risks associated with AI deployment.

• Accountability: Establishing clear responsibilities for decision-making and system performance.

• Data Governance: Managing the availability, usability, integrity, and security of the data used in AI products.

• IT Governance: Aligning IT strategy with business goals, ensuring efficient use of IT resources, and managing IT-related risks.

• AI Governance: Setting policies and frameworks to guide the development and use of AI technologies responsibly.

• Transparency: Making AI processes and decisions understandable and accessible to stakeholders.

• Privacy Compliance: Ensuring AI systems comply with data protection regulations and respect user privacy.

Business

Aligning AI capabilities with business objectives is a challenge I’ve faced repeatedly. This alignment is crucial for the product to deliver real value, specifically the value expected to be generated for the business. I focus on understanding how businesses invest in new initiatives, what capital costs are, and what it takes to assess the financial viability of our AI products. It also involves understanding market demands, competitor analysis, and strategic positioning. My ongoing learning in business strategy significantly influences how I approach product management.

These are some topics we need to consider when capturing the essence of Business:

• Business Models: Developing sustainable models for monetizing AI solutions.

• Strategy Alignment: Ensuring AI initiatives support overall business objectives.

• Performance Metrics: Measuring the impact and success of AI projects.

• Capital Investment: Understanding the financial requirements and securing funding for AI initiatives.

• Cost-Benefit Analysis: Evaluating the financial viability and potential ROI of AI products.

• Competitor Analysis: Assessing the competitive landscape to position AI products effectively.

• Sales and Marketing Strategies: Developing plans to promote and sell AI products.

• Project Management: Planning, organizing, and overseeing the AI product development process to ensure it is completed on time, within budget, and meets the defined goals and objectives.

People

It all starts and ends with people. AI products are built by people, for people. This element focuses on the team’s expertise, the user experience and needs, and the societal impact of our work. This is the most important element of the Double Trio. If we don’t understand what we are trying to solve, for whom we are trying to solve it, and with whom we are trying to solve it, we ultimately don’t solve anything at all and, hence, don’t build successful products. All the other elements are obsolete if we can’t navigate successfully through the most challenging aspect of AI product development: People.

These are some topics we need to consider when capturing the essence of People:

• User Experience (UX): Designing AI products that are intuitive and meet user needs.

• Team Collaboration: Fostering effective communication and cooperation among cross-functional teams.

• Stakeholder Engagement: Involving all relevant parties in the development and deployment process.

• User Research: Conducting thorough research to understand user needs, behaviors, and pain points.

• Change Management: Helping teams and users adapt to new AI technologies and processes.

• Leadership Development: Cultivating leadership skills within the team to guide AI projects effectively.

• Presentation and Persuasion: Clearly conveying ideas, project goals, and technical details to the team and other stakeholders. This includes presenting complex AI concepts in an understandable way, gaining buy-in from the team, and ensuring everyone is aligned with the project objectives.

You might have already recognized that we don’t live in a black-and-white world within the Double Trio framework. Some topics span multiple elements. For instance, AI Governance can fall under both AI and Governance, Presentation and Persuasion can be part of both People and Business, and MLOps can belong to both IT and AI. This overlap highlights the interconnectedness of these elements. Building these bridges requires someone to take responsibility. And you already know who my favorite is 🙂

Exactly, the AI Product Manager.

Balancing the Double Trio

Understanding the Double Trio Framework is about recognizing the interplay between these six elements. It’s about achieving a dynamic balance where technical capabilities are matched with ethical standards and market needs, and where people both shape and are shaped by AI technology. The real-world application of this framework involves continuous learning and adaptation, as each element can influence the others in substantial and often unpredictable ways.

To develop a successful AI product, it is essential to balance the Technical Trio and the Strategic Trio. This balance involves recognizing the interplay between these six dimensions and building bridges to address problems that span multiple dimensions. These bridges enable solutions for complex, overarching issues that cannot be resolved within a single dimension.

AI Product Managers play a crucial role in solving these complex, cross-dimensional problems by:

• Architecting Bridges: They create and manage connections between the technical and strategic elements, ensuring all dimensions are considered and integrated effectively.

• Balancing Elements: They continuously assess and adjust the balance between the Technical and Strategic Trios to adapt to changing needs and contexts.

• Active Problem Solving: They actively engage with teams and stakeholders to identify and address issues, leveraging their comprehensive understanding of both technical and strategic aspects to devise holistic solutions.

Example Problems and Solutions:

1. Data Privacy and User Trust:

   • Problem: Ensuring data privacy while maintaining user trust involves both technical and strategic elements. It requires robust IT infrastructure and data governance, along with clear business policies and transparent communication with users.

   • Solution: An AI Product Manager architects bridges between IT, data governance, and user engagement strategies to create solutions that secure data while fostering trust. They ensure compliance with regulations and communicate transparently with stakeholders.

2. Scalability and Business Alignment:

   • Problem: Scaling an AI solution to meet business growth requires seamless integration of IT and AI capabilities with strategic business planning and investment.

   • Solution: An AIPM balances technical scalability with business strategy by coordinating efforts between infrastructure teams and business units. They plan for scalable architectures that align with long-term business goals and secure necessary investments.

3. Ethical AI Deployment:

   • Problem: Deploying AI ethically involves navigating technical, governance, and people-related challenges. It requires developing fair algorithms, ensuring regulatory compliance, and considering societal impacts.

   • Solution: An AIPM builds bridges between AI development, governance frameworks, and user research. They advocate for ethical AI practices, ensure compliance, and engage with diverse stakeholders to understand societal impacts and address concerns.

💡 Once you find the right balance, you’ve found the golden cut, where an AI solution is ready to evolve into an AI product.

From there, AI Product Lifecycle Management begins, where we continuously balance the six elements to ensure the product remains viable. And you can imagine that this isn’t as easy as it might sound, since we need to continuously assess the ever-evolving landscape of technology and human needs, and adapt accordingly, finding again the right balance of the Double Trio to ensure we still have a viable AI product.

But this actually is a topic of its own.

I hope this framework will serve you well.

It definitely did for me.

JBK 🕊️

---

P.S. If you want to support me, for now, I am not accepting payments ☺️ However, you are welcome to share, like, and comment here or on my LinkedIn posts. This really helps me reach more people who are on the same journey as we are, and your feedback is invaluable in helping me improve the content iteratively. Thank you ❤️.

---