The Secret to GenAI Adoption: Five Small Steps, for Big Wins

In today's rapidly evolving artificial intelligence (AI) landscape, the sophistication of systems capable of processing human language and analyzing data is not just an academic discussion - it's a practical reality reshaping multiple industries. Generative AI (GenAI), with its ability to create new, contextually relevant content, is at the forefront of this transformation.

At ADAPTOVATE, our experience with GenAI is exemplified by a recent engagement with a prominent vocational education provider (details available in our case study) which resulted in staff being able to source information 10x faster, in addition to a 50% reduction in wait time for students to have their queries answered.

As organizations look to harness the benefits of GenAI and evolve their offerings, we provide five critical recommendations based on our experience, to ensure a successful implementation and maximize the advantages of this transformative technology.

1. Leverage existing IT infrastructure and data to deliver relevant responses while ensuring data governance

Public GenAI platforms are known to use any uploaded information to train their models and generate content. This is a security concern because it means that any private information submitted into the GenAI platform can be used by the platform to generate content for other users.

To address these security and privacy concerns, we recommend using the Microsoft Azure OpenAI service which leverages the capabilities of the OpenAI model while maintaining the security and enterprise features of Azure. This configuration stores all information within an organization’s existing Microsoft environment, while using local infrastructure that adheres to regional and organizational policies and regulations.

To further personalize and improve the accuracy and relevancy of GenAI answers, organizations can adopt a RAG (Retrieval Augmented Generation)¹ technique to augment the generative model's responses with accurate, context-specific information obtained from the organization's internal knowledge base. In our experience, using Microsoft Azure OpenAI, in addition to a RAG technique, ensures the organization can leverage its own data and the power of GenAI in a secure environment.

2. Prioritize small, easy to implement GenAI use cases and pilot them to realize and demonstrate early value

There may be initial apprehension to adopting GenAI inside an organization due to limited knowledge of GenAI’s capabilities, leading to difficulty delivering tangible benefits, uncertainty on how team members will react to using it in their day-to-day activities, or general resistance to change.

We recommend prioritizing simple and easy to implement use cases to start small and build from there. Do this by choosing use cases that deliver business value quickly but are also easy to implement and test during a short pilot (e.g. 2 weeks in duration). Starting small enables the organization to test the solution in a controlled environment and realize value quickly so that it doesn’t overwhelm team members, and it can build the business case for scaling GenAI across the organization.

3. Build the solution iteratively to ensure it addresses user needs, secures buy-in, and increases adoption

To allow for a GenAI solution to meet users’ needs and increase its adoption, it should be built through iterative cycles that quickly incorporate user feedback. The learnings from an initial pilot should be captured and used to enhance future iterations of the organization’s GenAI capabilities and help create a long-term roadmap.

This iterative process has three major benefits. First it enables the organization to realize business value sooner by launching the product as quickly as possible rather than waiting for it to be perfect, which delays benefits. Second, an iterative process will incorporate user feedback into the product, resulting in the product being fit for purpose and valuable for the end user. Lastly, an iterative process limits risk by enabling the team to pivot quickly if parts of the solution prove not to be valuable.

4. Run regular, short, and practical training sessions to effectively get team members familiar with the solution, and minimize disruptions to their work

Team members are busy with their daily responsibilities and often find it hard to engage in training sessions for new technology solutions. The increased mental load of learning to use a new technology, along with having different learning styles, can impact their current work performance.

Our experience tells us that minimizing the impact on team members’ daily responsibilities and reducing their cognitive overload is critical to learning how to leverage GenAI. As a result, we recommend short (30 min or less) and highly tailored training sessions focused on getting team members to work with the solution as quickly as possible. Practicing with the solution during training will improve team member skills through experimentation, while receiving guidance from experienced instructors.

Additionally, this training approach will generate opportunities for direct feedback, which can be quickly addressed by the implementation team, thereby removing barriers for team members to utilize the solution.

5. Capture and share the learnings from the pilot to scale the solution

To sustain the momentum from the implementation of the pilot and be able to build a roadmap to inform future direction, the learnings from the pilot must be captured and used to enhance the understanding of both the solution and the organization’s current GenAI capabilities. This will enable an organization to update the previously identified use cases, incorporate new ones that had not been considered before, and reprioritize them to create the roadmap.

This can be done by inviting teams from other parts of the organization to the learnings showcase sessions of the GenAI solution and involving them when identifying and prioritizing future use cases to build the roadmap. This will transform detractors to become early adopters and champions and foster a general appetite for innovation with proven technology.

Through the implementation and use of GenAI in our engagements we have been able to highlight its significant impacts, enhancing access to information and improving the user experience. By leveraging existing IT infrastructure, starting small and delivering iteratively, and running efficient, compact training sessions, large-scale uptake and true business benefits can be realized in any organization.

Footnote

1. 1 https://learn.microsoft.com/en-us/azure/search/retrieval-augmented-generation-overview