Taming dependencies, virtual team synergy, AI-driven product innovation | Dirk L?ssig, Digital Transformation Advisor, Thoughtworks
Yassine Kachchani
Co-founder & CEO at Gemography | Publisher of Exec Engineering, a weekly digest on Engineering + Talent
Today's guest is Dirk L?ssig , a Digital Transformation Advisor at 思特沃克软件技术（中国）有限公司 .
His 30+ years of experience in guiding companies through digital transformations provides invaluable insights. These perspectives are crucial for tackling the challenges of the current "AI transformation" organizations are undergoing worldwide.
Highlights
Yassine: You've written about managing ‘evil dependencies’ in software development. How do the challenges of dependency change with remote or distributed teams?
Dirk: The challenges remain consistent, but some become harder to address. My article concludes with: "The key is to align the organization towards value and to grow an inter-team collaboration culture." While the virtual nature of remote organizations might initially suggest that reforming and aligning teams is simpler, cultivating a thriving culture in such a setting is a significant challenge.
The first step in tackling dependencies is to align development teams toward value, which may require a new team setup. While assembling new teams might be easier in a remote setting, cultivating effective collaboration and trust within these distributed teams is more demanding. Of course, physical face-to-face communication remains the most effective way to build trust and foster psychological safety among team members. Therefore, distributed teams require more attention and care to achieve good collaboration.
While remote collaboration is now very common, resolving dependencies often involves tackling complex issues that demand effective communication between teams. This can be particularly challenging for individuals who haven't previously met in person, unlike in co-located settings where casual interactions, like spontaneous conversations over coffee, help build relationships.
Y: What strategies work best for addressing dependency in a remote setting?
Dirk: Regular cross-team collaborative events, even in a remote setting, are instrumental in fostering a shared vision and uniting teams around the overall product. In my experience, these events serve as a great booster for collaboration. Examples include big-room planning (often referred to as PI planning), vision or kick-off workshops, overall retrospectives, or unconferences. Even companies that primarily operate remotely often invest in occasional in-person gatherings to strengthen collaboration and team cohesion. These events have consistently proven to improve team effectiveness significantly.
Y: To what extent can the problems of dependency be proactively addressed through strategic hiring?
Dirk: It’s not something that can be fixed by a single strategic hire, as the ability of a development organization to manage dependencies is rooted in the capabilities and mindset of its people. However, leadership plays a crucial role in developing the organization and cultivating new skills and habits within the development teams. Look for a leader who understands all dimensions—business, technology, software architecture, and organizational dynamics.
Y: Are there specific traits that you think leaders should look for when hiring to anticipate dependency problems?
Dirk: Expertise in organizational design and culture building is crucial. While new technologies are making our organizations and their environments more complex, we remain stuck in a 20th-century mindset, thinking we can solve everything by simply rearranging the org chart and tweaking development processes. Navigating the complexities of the 21st century demands a new kind of organizational expertise, ideally distributed throughout the leadership team and across the organization.
Y: With many companies integrating AI into their core processes and products, what do you see as the biggest organizational and leadership challenges for tech leaders in this 'second wave' of digital transformation?
Dirk: The new opportunities brought by new technologies like AI are breathtaking. However, the mere technical feasibility of a solution doesn’t guarantee its desirability among customers, nor does it necessarily mean it makes sense for the business or is feasible within the organization. Exploring new opportunities must always start with the customer and be centered around the value it creates. Product development organizations must learn to "left-shift" AI expertise into the innovation process, incorporating AI considerations early and validating the feasibility of AI solutions upfront.
AI relies on data, and companies must prioritize their data transformation efforts before implementing AI solutions. Data Mesh offers a valuable socio-technical approach to building a strong data foundation. While this transformation is a significant undertaking, it ensures the accessibility and quality of data needed to unlock its full value within the enterprise.
Another challenge lies in deploying AI solutions into production. Many companies struggle to move their AI projects beyond the proof-of-concept stage. To overcome this, they need to adapt their development processes—such as CI/CD and validation—to effectively deploy AI solutions in live environments.
Y: What steps should tech leaders take to prepare their teams—both culturally and technically—for this change?
Dirk: Tech leaders should start by defining a data strategy, followed by an AI strategy. These strategies should include Data & AI governance, covering the ethical implications of AI as well. Implementing these strategies will require changes to the operating model and upskilling within the organization, best achieved through a thin-sliced approach. Additionally, leaders should proactively build a portfolio of partners to strengthen their internal capabilities and stay ahead of the curve.
Y: How should tech leaders approach talent acquisition and team composition while adapting to AI and similar changes?
Dirk: Specialized AI talent, particularly in areas like large language models (LLM) or machine learning, is scarce, making it difficult for companies to upskill their teams. Therefore, I recommend collaborating with partners like Thoughtworks to explore potential opportunities before committing to hiring your experts. The skills you need to hire will depend on your AI strategy and the specific use cases with AI potential. Working with partners helps maintain flexibility. Regardless, the AI experts need to be embedded in the development teams responsible for delivering those AI use cases.
It's still a smart move to upskill employees on general AI topics and introduce AI-powered software engineering tools like Copilot to the development teams. These tools can boost productivity throughout the entire development process, from initial product discovery to operations.
Y: How do you envision the ideal software development organization of the future, and what steps can leaders take today to move towards that vision?
Dirk: Currently, there is a trend toward establishing organizations with independent, autonomous teams. However, my article suggests that this approach has its limitations. Value isn't confined to the focus areas of individual teams; it also emerges through collaboration among them. To create superior solutions that truly benefit customers, we must transcend team boundaries. In my ideal organization, dependencies between teams would be seen as opportunities for collaboration, not obstacles.
Furthermore, the growing use of AI-powered software engineering tools for code navigation and generation will reshape software development. Currently, scaling beyond team boundaries is often hindered by the cognitive overload of navigating extensive codebases. These tools will help mitigate this challenge, enabling teams to implement features more rapidly by simplifying the process of understanding and modifying unfamiliar code. As a result, dependencies on other teams will decrease as teams gain the autonomy to make necessary changes themselves.
Thank you Dirk for your time and insights!
This interview is part of the “Exec Engineering Dialog” series where I interview seasoned tech leaders on the topics of talent, product, management and culture.
If you liked the insights shared in this interview, consider giving feedback and/or sharing it with your network, it’s the best way to help this segment improve and grow.
P.S. If you prefer your content on Substack, I'm also there.
Yassine.