Ravi Jasti, Founder and CTO of AirQuery and a data technology enthusiast.
We should be looking at AI as the fuel that can propel Business Intelligence’s ability to problem-solve.
Jane’s Case
Let’s consider a hypothetical and walk through the use cases and potential pitfalls with AI-augmented Business Intelligence
Jane’s situation: She’s a high-performing sales professional. She routinely has to go through dozens of operational reports, performance reviews and other strategy plans from many branches. Jane has to meet one of her key clients in the next 10 minutes today.
For this meeting, she has to access a sales report and structure her pitch based on details of a prior business, this customer and other comparator data. She has to get this information immediately, the data needs to be correct and she needs to understand it properly—then glean insights that will directly help her in her current task.
Jane’s Journey
In this scenario, Jane’s data location journey would involve various stages and several obstacles as well:
Data Sourcing
Jane must gather data from various sources, such as databases, data warehouses, or external data providers. She must first identify the sources from which the customer data will be collected.
Data Extraction
Then, she must extract the relevant customer data from those sources. In some cases where people look for technical or niche data, this could involve querying databases, running scripts or utilizing extraction tools to retrieve the required information. Companies usually rely on a separate team to give you that information.
Data Transformation
After extraction, Jane may need to transform this data to ensure it is stable for quick reference or ad hoc reporting later. This could include cleaning the data, filtering out irrelevant information, aggregating or summarizing data, or performing calculations and derivations.
Data Storage / Temporary Location
During ad hoc reporting, Jane must temporarily store or copy this to a specific location for analysis and processing. This could be a local filesystem, a cloud storage service or a dedicated ad-hoc reporting environment.
Data Analysis
Jane will also have to perform exploratory data analysis, generate reports, create visualizations or run statistical models to derive insights from the client data. This could involve using tools like spreadsheets, business intelligence software or ad-hoc reporting platforms.
Data Presentation
After the analysis is complete, Jane will need to present the results of the ad-hoc reporting in the form of reports, dashboards, charts or presentations. Her pitch will be composed of these insights. These outputs can provide valuable information about customers.
Data Retention/Archiving
Depending on Jane’s company’s data retention policies and regulatory requirements, her collected data may need to be retained or archived for future reference or compliance purposes.
Challenges Faced And Problems To Avoid
Data Overload
Jane will have to sift through large volumes of information because her company has acquired a lot of new customers in the past.
Strain On Time And Efficiency
Typically, Jane will need to spend much time getting the information she needs. More than 15 minutes will be required.
Data Visualization Is Challenging
Although Jane would like to outsource the task of making a pitch deck to someone else, they may not have the skill set to do it.
Data Quality Is Not Good
After Jane gets the data, the raw data may need cleaning up, the information itself may be missing or outdated, etc.
Assistance With AI
Another BI solution may not really help with Jane’s tasks. She needs a solution that can streamline, simplify and automate her work.
Reporting With The Help Of Natural Language Queries (NLQ)
This eliminates the need for complex coding and makes data exploration more accessible.
Provide Intelligent Insights
AI-powered BI platforms can facilitate faster and more accurate discovery of valuable insights. This can help them make more informed decisions.
Be Incorporated Into The Existing Tools And Technology
This means that users like Jane do not have to spend time and effort learning new software.
Design Approach To Infuse AI Into BI
Empathizing With The User And Plotting The User Journey
Factor in the requirements of business users for high-level information and specific details. For example, a salesperson must know about the range of coffees available in a store. They will also need to know how many hazelnut-flavored bottles were sold last May in one store.
Providing Contextual, Augmented Results
Use both metadata and real data to provide augmented results. For example, providing metadata of all items in a store along with real data on the sales promotions that ran for the coffee stock in a particular store can help predict demand.
Provide A Smart Search
Feature that can be integrated into any business application and you can search for details within whichever application you use. For example, if a salesperson wants information on a customer, Search for contextual insights from within the document, email or Excel sheet the user already uses. There is no need to access another application, etc.
Exploring And Creating Your Data Routes
Allow users to explore and find their own data trails. For example, if users do not find what they are looking for, they can explore different aspects of the information on the subject and create a personal guide of the information the user needs.
Potential can now be realized through a simple click, whether in data or through people.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
Read the full article here
