The Path to the AI Summit
As Edmund Hillary and Tenzing Norgay prepared for their final summit push, they thought about and planned their activities for the next day. They were perched in a single tiny tent at Camp IX at 27,900 ft above sea level. To get to this camp, Hillary had to carry 63lbs of equipment and Norgay 50lbs as they used a long-handled ax to carve out steps on the narrow ledge with steep drops on both sides. As night fell, they focused on sustaining their energy and body heat as they ate sardines on biscuits and tinned apricots. Setting off to climb the Mt. Everest was one thing; sustaining their energy at this elevation and planning their final push was another.
Our point of view is that reaching the AI summit requires climbing three distinct mountains, the Setup, Sustain, and Scale mountains. Like climbing high mountains, the Sustain phase of the AI mountain requires a different set of activities to be performed and skills to be utilized. This is where the AI teams must work side-by-side with business teams to implement business processes changes and drive measurable outcomes.
The Second Mountain: Sustaining Artificial Intelligence
During the Sustain phase, the goal of the organization is to unlock and capture the business value from the AI recommendations. Until this point, the value of the AI team has been theoretical and includes recommendations for the business to implement. Additionally, to sustain the momentum, the AI team will need to begin planning and working on longer-term AI investments.
In our eBook, we explore the five steps to ensure you can sustain your AI momentum. These include optimizing your business processes, operationalizing solutions and measure the impact, prioritizing clean data initiatives, developing an insights-driven culture, and finally creating an AI strategy.
Optimize Business Processes
By using AI models, the team will be able to generate ideas and recommendations about change that can drive business value. For example, we may have found a way to comb through medical records to recognize a certain negative health outcome and we may want the doctor to change their treatment, or we may have found that a certain vibration pattern and temperature gradient that when combined increases the likelihood of failure of field equipment and we want to make a change to proactively mitigate the risk. These modeling recommendations often lead to changes that need to be integrated into operational business processes. This is where the rubber meets the road. This is where the climber now climbs the rockface. With the theoretical knowledge, the team now puts this skill into practice
Operationalize AI Solutions and Measure Impact
In addition to making the needed business process changes, the technology might also need to change. For example, the data that was used for the initial algorithm might have been sourced from multiple systems. For the algorithm to now operate in production, it might need substantial technology enhancements to support the data integration between the various systems. Additionally, the algorithm might need to be moved to a support team to ensure it can be maintained and run periodically to continue measuring the impact and outcomes. Unfortunately, without the right planning and structure, AI teams may end up developing and supporting their models. Over time, the support related tasks continue to increase, reducing the amount of new development the team can then take on.
- Prioritize Clean Data Initiatives
Climbers not equipped with the right crampons attached to their boots will find themselves wasting energy with each step on the glacier, making the climb to the summit extremely difficult. Like the climber, the AI team not equipped with the right data foundation will find themselves spending an inordinate amount of time prepping data. By most estimates, data scientists not set up with clean data platforms spend upwards of 80% of their time getting data prepped for modeling and analysis.
- Develop an Insights-Driven Culture
Climbers pay special attention to the weather. They know that to summit successfully, the environmental factors must support them. In a similar fashion, the AI team will only be successful if the organization sees value in AI and leans on insight-driven decisions. AI focuses heavily on data and insights and the organization must be comfortable with this. Developing an insight-driven culture tends to be easier for newer organizations and requires significant effort and investment for older organizations.
- Create an AI Strategy
Climbers don’t put together their plan for the big mountain without first having successfully attempted some of the smaller mountains where they have built up their endurance, capabilities, and confidence. Similarly, AI teams should develop their AI strategy only once they know their capabilities. This helps develop a better-informed plan and increases the momentum to drive the change through the organization from executive funding to front-line personnel involvement. It provides a way to take this from AI strategy to AI execution.
Why Climb the Mountain?
Your efforts to sustain your AI investments will require you to lean into your business partners and ensure that the organization is incorporating the recommendations, changing business processes, increasing data literacy and adopting an insight-driven culture, scaling up the technology to support production use, and focusing on the longer-term plan to take your AI investments to the next level.
Whether you are at the start of this journey or somewhere along the way and looking for a different route to the summit, we welcome you to reach out to us at Sense Corp. To learn more, download our eBook, Reaching the AI Summit: The Definitive Step-By-Step Playbook for Setting Up Your AI Organization.
Sense Corp offers a variety of consulting services across data, digital, and business transformation. Our team of skilled consultants offer insights and solutions to the most important problems facing business and government. To learn more about the services we offer, visit our website or contact us.