Image by authorWhat would happen if AI didn't exist? In some ways, there has never been such a technology frenzy sweeping through the industry.
For business leaders whose central focus is leveraging technology to drive business growth, the first thing they think about is their customers: who are they serving? Who is our audience? What do they want from us?
And what immediately comes to mind is their problem. What do they need that not even their competitors are offering?
Customer-focused business strategy
And that begins a series of questions that, if addressed, can lead to a successful business.
- What makes your customers' lives easier?
- What will make the experience seamless for them?
- What are their sufficient needs?
And so begins the path to discovering the means to an end: technology.
In particular, AI has not yet been discussed. Listing your business strategies, tools, and privileges is the most important and most important step in determining what you are solving and who you are solving it for.
after that. The question arises, “How to solve this?” Could AI be a good solution to this business problem?
Today, enterprises need a framework to determine which use cases AI is suitable for. This is what I propose, the “PRS” framework.It stands for “”pattern that repeat in scale”.
Image by authorpattern
Let's look at an example of internalizing this framework.
For example, taxi service providers offer taxi driver availability at cost-effective prices, taking into account various factors.
- Available driver pool and taxi requestor distance
- distance to destination
- There are more people requesting taxis than there are taxi drivers, so peak demand leads to higher prices.
- Reportedly, a taxi requester's phone battery is running low, which could signal a fare increase. This gives the taxi service provider a signal that when the mobile phone battery is low, the taxi requester may feel a sense of urgency and be more willing to pay more for the same trip. can do.
- Taxi availability and prices also vary depending on factors such as regular vs. premium taxi service, time of day, and inclement weather.
Image by author
By doing all this, taxi drivers will have enough incentive to continue to enhance the customer experience.
Therefore, you can understand patterns in your data.
large scale repetition
Next, reproducibility is important. All of these data attributes are repeated for each taxi requester and each geographic trip, which inevitably leads to the final point: scale.
Consider how unattainable this problem would be if there was a manual or non-AI workflow to solve this compute-intensive business case.
data strategy
Once you've developed your business mindset and identified the problems that are well-suited to being solved by AI, it's time to turn all your attention to data. After all, data is the core engine that drives the success of all AI algorithms.
We also have a framework for this called AAA, which stands for Availability, Accessibility, and Authorization.
Consider this:
Do you have data?
versus
Is there comprehensive data?
There is a small but crucial difference between these two statements.
Just having data is not enough. For a model to be able to recognize all the attributes that a human expert can recognize, it needs all the data needed to model the phenomenon. Therefore, data availability is important.
Next is data accessibility. Having data at your disposal is one thing; having data easily accessible is another. Building data pipelines is important to ensure seamless data access.
So far, you've done a lot of work to shape your data, but what if you're not allowed to use the data for model training or analysis purposes?
This is where most organizations stumble. Make sure you have the necessary permissions, or better yet, only use data that has the necessary permissions.
There is one unanswered question regarding the 3A's of data strategy. So, what is the sequence or sequence between business strategy, data strategy, and AI strategy?
There are lots of strategies!
Generally speaking, AI strategy is always aligned with business strategy and aligned with data strategy. While maintaining the 3A’s of data in progress, it is wise to continue working on AI use cases as well.
Similar to the iterative nature of AI projects, AI roadmaps must be continually refined while preparing and hardening data infrastructure to maximize the potential of AI technology within an organization.
Continue to analyze and track key performance indicators (KPIs) such as accuracy, efficiency, and ROI to regularly assess the status of your AI initiatives to assess their effectiveness and identify areas for improvement. Masu.
bonus chips
Most AI projects and strategies are affected by lack of timely communication. It's important to perform milestone checks and actively seek feedback from stakeholders such as end users and business leaders. Every successful AI project goes through several cycles of iteration, with feedback informing adjustments and enhancements to existing models or the development of new use cases.
Furthermore, models are developed once and never revisited. Your business priorities are likely to change over time, and your AI strategy and implementation should reflect that.
Vidhi Chu He is an AI strategist and digital transformation leader building scalable machine learning systems at the intersection of product, science, and engineering. She is an award-winning innovation leader, author, and international speaker. She's on a mission to democratize machine learning and break through the jargon so everyone can be a part of this transformation.
