Operating models for gen AI products need to accommodate changes in a variety of areas, including data practices, development and engineering skills, governance, cost management, and benefits realization. Each change will create new handoffs and as an operating model matures, the handoffs will continue to change. For example, how should a gen AI platform team hand off work to a data ingestion team, and vice versa? Organizations using an ad hoc “by-default” approach will struggle to sort out which handoffs need to change, which should be retained, and which can be eliminated. Hybrid by design adds clarity and focus to the way forward by providing a well-tested operational foundation.

Remember that most handoffs come from organizing work around specialized, functional silos. Designing gen AI operating models “by default” can make the handoff problem worse if every new element of gen AI tech gets organized around new silos: platform teams, ingestion teams, inferencing teams, and so on. Here are more recommendations for attacking the silos that create 78-handoff operating models:

Map the money trail. Imagine a clear illustration, a value stream map, that tracks every dollar spent and every handoff that occurs from the moment a product idea is conceived to the moment cash lands in your accounts. This map will expose the bottlenecks and redundancies that are currently acting like financial sinkholes, draining both time and money. In other words, before you can fix the problem, you need to understand it. Visually map the handoffs within your product development cycle, from concept to cash.

Edit ruthlessly. Not all handoffs are created equal. Analyze your map and make the tough calls. What can be eliminated? What can be automated? Where can processes be streamlined? Focus on a single, high-impact handover (or decision point) that you can tackle immediately. Fix it this week for this program and use it as the start of a virtuous cycle that repeats itself.

Unleash the hackathon heroes. Light a fire under your teams with a handover hackathon. Challenge them to slash the number of handoffs within a project. This gamified approach will not only generate creative solutions, but also foster a culture of ownership and efficiency. Remember, the goal is to go from a plethora of handoffs to a single, streamlined flow. Incentivize participation. Reward teams for their ingenuity and showcase the improvements across the organization. This positive reinforcement loop will turn your initial win into a virtuous cycle of continuous improvement.

The idea that business leaders and tech leaders aren’t speaking the same language isn’t new and it hasn’t gone away. During the rush to cloud it was hard enough for tech leaders to engage their business unit counterparts in discussions about containers, landing zones, and cloud-native development. Now business leaders and tech leaders alike need to learn the language of gen AI: prompt engineering, data ingestion, model training, and more.

The work of designing gen AI operating models runs the risk of driving both parties into their separate corners, but hybrid by design can make it an opportunity to bring them together in deeper collaboration.

Silos stifle potential. Collaboration is the key to unlocking the next level of performance in any operating model—it greases the wheels. By harnessing collective intelligence and expertise across the organization, enterprises can streamline processes, accelerate innovation, and ultimately drive superior results. Here are some strategies to break down barriers and foster a truly collaborative environment, using hybrid by design to drive exponential innovation.

Use the real world as your learning laboratory.
Teams bond and thrive on real-world experimentation. Hybrid by design incorporates agile methodologies, such as sprints and scrums, to facilitate rapid iteration and experimentation in real-world situations. This approach enables teams to respond quickly to changing requirements, iterate on their designs, and deliver working solutions. Build this type of learning and collaboration into the operating model. As you do, collaboration becomes fun and produces better results, eventually becoming the by-product of shared success rather than an aspirational goal.

Celebrate failures.
Fear of failure can be palpable. Mistakes are inevitable. To mitigate potential inaction, foster an environment of open communication where blame games are replaced by problem-solving. Failures that are shared and used as learning opportunities are actually wins. Leaders cannot encourage experimentation and cocreation between the business and IT if they punish failures with poor reviews or worse. Ask teams to share learnings gleaned from failed attempts as well as successes. Both are learning opportunities for other teams. Gen AI can help. Learnings often get siloed within teams or individuals. Train a gen AI assistant to identify key takeaways from various sources such as team reports, project discussions, or internal communication platforms. Set up a regular schedule where the AI assistant can automatically generate summaries of these learnings. This keeps the information fresh and readily available for other teams.

Build a skills swap program.
Break down silos through strategic role rotations. Rotational redeployment involves strategically moving employees from one role or department to another within the organization for a set period. This cross-pollination fosters empathy for each other’s challenges and creates a shared understanding of the bigger picture. Start by assessing skill gaps across departments and long-term project needs. Where could a fresh perspective or a specific skillset be beneficial? Look for adaptable, high-performing employees with a willingness to learn new things. Offer training opportunities tailored to the new role, including soft skills such as communication and collaboration across departments.

Hybrid by design focuses on two main shifts in operating model design. The first is the ongoing shift from “by default” hybrid cloud adoption to more intentional, business-focused hybrid-by-design investments.

The second is hybrid cloud’s role as it becomes the foundation for building gen AI products. Most organizations are going through both shifts at the same time, and leadership teams must get very focused on designing and documenting the changes required, as well as the game plan for executing those changes. But it’s easy to overlook what we’ve all experienced directly: the difference between plans that work on paper and plans that work in the real world lies in teams’ level of engagement. Getting engagement right can’t be limited to pro forma engagement surveys—it needs to be reflected in managers’ daily practices.

Here are a few key ways to put the focus on people so your operating model can hum like a well-oiled machine.

Equip employees for decentralized decision-making.
By giving employees more autonomy and access to cloud-based tools, hybrid-by-design models enable decentralized decision-making, reducing the need for hierarchical approvals and increasing speed and agility. Cloud-based project management, collaboration, communication, and data analysis tools empower employees with what they need to make faster, better-informed decisions.

Incorporate anonymous feedback into AI models.
By incorporating sentiment and engagement data from various sources (for example, employee surveys and feedback, workplace metrics, and so on) organizations can create more accurate and nuanced predictions of employee behavior, engagement, and performance. This allows them to address problem areas before they become severe, and in some cases, prevent the problem entirely. Hybrid by design allows the capacity and flexibility to do this in real time, enabling AI to deliver.

Track what no one else is tracking.
What’s your idea leakage rate? This is the percentage of employee-generated ideas that get lost or abandoned somewhere in the approval process. This could indicate a lack of clear pathways for ideas to travel, or a culture where employee suggestions aren’t valued. How about your fix-it-myself rate? Pay attention to how often employees bypass official channels and solve problems on their own, even if it takes more time or effort. A high fix-it-myself rate could signal frustration with the existing approval process or a lack of faith in getting timely solutions from above. And finally, how often do you hear the words, “Yes, but…”? Monitor the frequency with which new ideas are met with initial resistance or negativity (“Yes, but we tried that before...” or “Yes, but it won’t work because of...”). A high “Yes, but...” ratio could indicate a culture of risk aversion or a lack of psychological safety for employees to propose innovative ideas. Here again, hybrid by design allows the capacity to track an almost infinite number of factors, while AI provides speedy, integrated analysis.

Gen-AI-infused operating models can unlock innovation. But it takes people to deliver that innovation. Enterprises are approaching hybrid-by-design operating models in ways that work for them, but no two paths are the same.

Hybrid by design in action