Data analysis is moving from the BI era to the AI era: Three reasons for accelerating the democratization of data
Data analysis is moving from the BI era to the AI era. Even if you have taken the trouble to introduce BI tools, are you worried about "democratizing data" because only a limited number of people can use them, or because analysis speed is slow?
In this column, we explain the new standard for data utilization using generative AI, which we have arrived at after 10 years of trial and error. We also introduce three compelling reasons why we should evolve from BI to AI, where anyone can instantly obtain analytical results simply by asking questions in natural language (the language spoken by humans). We also introduce tips that will dramatically change your organization.
A decade of progress towards generative AI
Since ChatGPT began to attract attention a few years ago, the word AI has become so prevalent in the news that it's hard to go a day without seeing or hearing about it. Many of you readers are likely using generative AI in various aspects of your business. Recently, a system called RAG has become increasingly common, and efforts to utilize data accumulated within companies with generative AI are on the rise.
Our company has been working on various data analysis and utilization initiatives for the past 10 years. The main flow of these efforts can be summarized as follows:
- Step 1. Management dashboard development (visualization of business status)
- Step 2. Company-wide deployment of dashboards (implementing necessary visualization in each department)
- Step 3. Build a data platform (employees can use data whenever they want)
- Step 4. Use of generative AI (using data while conversing in Japanese)
*We will not go into details in this column, but if you are interested, please feel free to contact us.
▼I want to know more about generative AI
⇒ Generative AI | Glossary
The barrier to "data democratization" that BI alone could not overcome
The key point I want to focus on here is the transition from Step 3 to Step 4.
In Step 3, we collected all data within the company, allowing employees to access it whenever they wanted and create their own dashboards using BI tools, which led to a rapid expansion of data analysis and utilization. However, even though we are an IT company, there were still a certain number of employees who could not use BI tools effectively or access data, and we still had challenges in advancing the "democratization of data" that would make us a so-called "data-driven company."
How to use generated AI data that "anyone" can analyze in Japanese
To solve this problem, we decided to utilize generative AI in Step 4. Since internal data had already been collected, we upgraded the platform by integrating generative AI into it. This created a system where questions asked in Japanese would be answered based on internal data. The screen image is shown below.
The screen image above captures the moment when the AI responds to the question, "Tell me the sales figures for the last three months." Of course, it also responds to additional requests, such as "narrow it down to product A" or "show me a graph," allowing users to interact with the AI to find the data they want, understand the situation, and dig deeper to take action.
When we introduce our company's initiative, many companies express interest and say they would love to try it. Although it is not yet widespread across the world, an increasing number of companies are now using generative AI to conduct data analysis.
Why is data analysis moving from BI to AI now?
Until recently, data analysis typically involved using BI tools (such as Tableau or Power BI), but why has generative AI become the new trend now? There are three main reasons.
- Low threshold and high level
- Overwhelmingly fast speed
- Reduced license costs
Now, let me explain each one.
① Low threshold and high level
Some of you may think the title is contradictory (lol), but let me explain. First of all, with data analysis using generative AI, users basically don't need to be aware of where the data is located, and they don't need to know how to read a dashboard; as long as they can speak Japanese and type on a keyboard, they can perform the analysis. In that sense, I think it's very easy to get started.
On the other hand, the questions (prompts) you ask are not just one-time questions, but you can dig deeper into any points of interest, allowing for a higher level of analysis.
for example
- What percentage of total sales are new?
- What usually triggers new business negotiations?
- How long does it take to close a deal?
- How much does this timeframe vary depending on the salesperson?
- What measures can be taken to shorten the period?
With the right data, you can dig deeper and find more specific insights.
In that sense, I think it's a particularly good tool for managers (such as relatively high-ranking managers) who want to focus on making important business decisions rather than spending time on specialized IT operations.
② Extremely fast speed
In the previous section, we listed some examples of deep data digging, but it will likely take a certain amount of time (one to several months) to visualize these using a BI tool (of course, this will depend on the circumstances of each company). If you ask your company's information systems department or an IT vendor to do this, it will take time and cost money each time.
If you have a set of data and perspectives you want to see, and you want to view them regularly, it may be better to have the information systems department or an IT vendor create it, but if you want to "try it out" or "for example," or you're in the process of trial and error, such as wondering what would happen if you looked at it this way, I think a generative AI environment where you can ask questions and get an immediate response is more effective.
3) Reduced license costs
So far, we've written about the benefits of data analysis using generative AI, but finally, let's look at the cost. One of the complaints we often hear from customers is that BI tool licenses are expensive. Of course, the cost varies depending on the tool, but as the number of users increases, the cost inevitably increases exponentially, so many people are probably confused when asked, "Do we really need one for everyone?"
When we build a data analysis environment using generative AI, licenses do not increase with the number of users, so we believe it is possible to expand data analysis throughout the company without worrying about license costs. We hope you will take this opportunity to consider it.
Taking data democratization to the next level
So far, I have explained "Data analysis is moving from the BI era to the AI era - Three reasons why data democratization will accelerate." Did you understand?
Those who are already working with AI may be wondering, "Won't generative AI have problems with the accuracy of its answers?" Meanwhile, those who haven't yet started using it may be worried, "Is it okay to create an environment where generative AI can view all of the company's data?" We offer a variety of solutions to answer these common questions. If you're interested or have any concerns, please feel free to contact us.
Saison Technology Online Consultation
If you would like to hear more about our data utilization platform, we also offer online consultations.
The person who wrote the article
Recommended Content
-
Generative AI opens up the "democratization of data"
Through our use case studies of generative AI, we will explore specific approaches to breaking through the "two barriers" that prevent all employees from utilizing data. -
BI Tools vs. Generative AI? How to Use Data Together in the AI Era
We will organize the characteristics of BI tools and generative AI, which are currently attracting the most attention in the field of data utilization, and how to best use them. -
Democratizing data analysis with generative AI
We will introduce a method that uses generative AI to enable anyone, including business users, to extract value from data.
