Making sense of enterprise data can be challenging for organizations of all sizes and sectors. Not only do organizations often process too much data more than they can manage, but they also struggle to extract value from the data they store.
The prevalence of data-blind practices means that most data is left untapped. According to a study by Seagate, only a third of data is ever used before it is created, leaving as much as two-thirds of their data going untapped in a world where data is crucial to business success.
So how can organisations unlock this hidden data?
Generative BI (Gen BI) by Pyramid Analytics is a solution designed to solve this problem, providing data teams with the business intelligence they need to make the most of their data.
This article tells you everything you need to know about Gen BI by Pyramid Analytics, including what it is, some of its key features and its use cases.
What is Gen BI?
Generative BI (Gen BI) is a business intelligence (BI) tool that leverages AI to transform the way businesses interact with data. Introduced in March 2024 by Pyramid Analytics, it combines the capabilities of AI with traditional BI tools to give users actionable insights quickly and seamlessly.
Whether it’s creating dashboards, finding explanations for data anomalies, or forecasting data trends, Gen BI automates crucial data management tasks.
Thanks to the tool’s natural language processing (NLP) capabilities, users can ask it questions in natural language and receive accurate and easy-to-understand responses in seconds. The tool will automatically analyze the data, identify relevant trends, and provide clear and concise answers to help the user understand and resolve their data management issues.
These NLP capabilities are paired with Pyramid’s PYRANA analytical engine for conversational analytics, which uses a semantic model to generate SQL queries that analyze data automatically. When users ask questions or prompt Gen BI to create reports, the semantic model makes it possible to visualize data through the Pyramid user interface. This enables data teams to perform data analysis at scale without limitations, directly from the data source.
Gen BI can also figure out and automatically perform all the necessary analytical steps to deliver any requested insight, meaning users don’t need to worry about analytics. This level of automation eliminates the need for extensive data preparation and analysis, enabling businesses to make well-informed data-driven decisions effectively without the learning curve.
In essence, Gen BI builds upon the capabilities of Gen AI but adds a layer of automation and user-friendliness, making it a more accessible and efficient tool for data analysis.
How does Gen BI work?
The smart business intelligence chatbot powered by Gen BI works as an intelligent assistant capable of understanding and responding to natural language inquiries about data.
Users can interact with Pyramid’s chatbot via text prompts or speech to ask questions about analytics processing. They can also request the tool to create dashboards, reports and presentations, using prompts such as “filter”, “sort”, “add” or “remove dimensions” to drill down into the underlying data and change the visualization for the best approach to analyzing complex data.
Pyramid CTO and co-founder Avi Perez explains that Gen AI merges the company’s powerful analytics engine with a multi-LLM AI solution to provide companies with a highly adaptable decision-making solution.
“Pyramid is melding the best of LLMs (Large Language Models) with its powerful analytics engine to usher in the Generative BI era to provide companies with a, and highly adoptable decision-making solution,” Perez said in a statement announcing the tool.
“Generative BI will transform working habits and how data is analyzed, enabling anyone in an organization to access and explore data, and ultimately get the answers they need quickly.”
Gen BI is especially useful for business leaders looking at straightforward answers from complex data in just a few seconds. It can also provide explanations on vague or relative terms, leaving room for more insights at a shorter duration.
Users can make complex multi-construct requests in any language and further examine the results through follow-up requests in addition to the analysis already presented by the BI bot.
They simply need to relay to the bot in natural language to either add more criteria or amend the visible data for a satisfactory tailored dashboard, allowing them to craft highly formatted reports in a matter of minutes.
Key features of Gen BI
1. Multi-LLM Framework
Gen BI caters to the Gen AI of today and tomorrow through its “Multi-LLM” strategy framework. The tool can customize data visualizations and analysis for certain tasks and domains to ensure that the most appropriate LLM is used for each query, leading to more accurate, custom, and relevant results.
LLM consolidation also allows users to gain qualitative responses in natural language, mimicking humans in their answers. Similar to other LLMs such as OpenAI’s GPT family and Google’s Gemini, the tool allows a natural conversation and can answer questions on a range of different topics. The difference is that Pyramid’s Gen BI is specifically presenting a solution for data analytics business intelligence.
This makes Gen BImore a specialized application of Gen AI specifically designed to automate the data analysis and management process by automatically handling tasks such as database connection, recipe generation, and query execution. This automation allows users to focus on asking questions and interpreting results, rather than worrying about the technical details that often come with your standard LLMs.
2. Answering complex questions
The gen AI-powered tool manages complex analysis via conversational natural speech putting the users at ease. This approach democratizes data analysis by making it accessible to users without extensive technical expertise.
“You need a real-world, conversational analytical companion to solve complicated business problems,” Pyramid Analytics says. Gen BI handles complex analytical questions and scenarios that a user can literally deliver through speech. This allows the user to quickly prep data to publish it through speech prompts, generating simplified analysis, and developing multi-layered content while dramatically speeding up decision-making.
Users can also create pixel-perfect reports in about 2 minutes to design a high-quality and visually appealing PDF, Word, PowerPoint slide, or a PNG image file for the data-backed reports.
3. Compatible with data types
Functioning with various types of data, Gen BI can work directly off any data source, eradicating the need for considerable data preparation or ETL (extract, transform, load) computer processes. This increases accessibility and efficiency for businesses with diverse data types and users can seamlessly connect to any data source including Redshift, Snowflake, Databricks, Microsoft Fabric, or SAP, and support both SQL and MDX.
4. Security by default
Pyramid Analytics’ Gen BI tool prioritizes a privacy-preserving approach when users present data – sensitive or not. To ensure data security and compliance, the firm avoids the transmission of private data to LLMs, ensuring private enterprise data remains confidential.
Instead, It passes the information that helps the LLM understand your data context, like schemas and semantic models.
Pyramid says: “You’re not limiting your data access when connecting directly to the data source, your confidential data remains preserved, and the business logic remains intact without compromising on security or governance.”
“Everybody wins in this situation.”
5. Seamless integrations
Gen BI can integrate into various common enterprise applications, including Redshift, Snowflake, Databricks, Microsoft Fabric, and SAP. The tool also effortlessly merges AI-driven analytics into existing applications and provides users with intelligent, conversational analytics directly within the software they are using.
Real-world examples of Gen BI in action
1. Data analysis
Gen BI can give users quick insights into data by automating multi-construct requests in any language. Users can follow-up requests and interrogate the tool for additional analysis, and the data can be discovered through conversations with the AI-powered chatbot.
For example, a conversation with Gen BI could look like:
User: “Compare sales to product returns across state country and show regression of sale to returns.”
Gen BI: “The regression analysis indicates a strong relationship between product returns and sales across different states and countries,” AI added in a few seconds.
User: “Explain what’s going on in California.”
Gen BI: “In California, Sales are predominantly driven by the Bike’s product category which represents 92.7% of sales, with the ‘U’ style accounting for 93.1% of Bike’s category and 86.2% overall while the ‘Reseller’ promotion application contributes to 77.8% of the ‘U’ style sales and 67.1% overall.”
2. Prompting Gen BI to create dashboards
Pyramid’s natural language interface and LLM framework enable users to design storyboards using Gen BI. With both text and speech, users can make complex multi-construct requests in any language.
For example: this could look something like this:
User: “A sales analysis slide showing me quantities sold by occupation and education. A different comparison of returns and customer happiness by state and last, an analysis of country sales performance by promotion using revenues, expenses, and margins. And I want it sliced by years with slide insights.”
Gen BI (presents a dashboard on the screen): “This dashboard provides an analysis of sales quantities by occupation and education compares returns and customer happiness by state and evaluates country sales performance by promotion in terms of revenues, expenses, and margins.”
The user can use Pyramid’s layout tool to adjust the design as required.
3. Formatting reports in minutes
Gen BI helps users build detailed reports using multiple queries across multiple pages without lifting a finger. Users simply need to select the publish button on the left bottom corner of Pyramid’s system, choose a theme and then hit the create button then pick a data model.
The chatbot automatically asks the user to pick the data model when you hit create:
User: “I’d like you to build me a report with the following sets of analytics. I want to see a comparison of sales by age and gender. I want to see a different report showing me the profitability by product, subcategory, and style. I want to see sales in all months. I wanna see an analysis of country and state sales performance. And last, I think I want to see a chart showing me manufacturer performance with their groupings, showing me returns, sales, and expenses, and I want it sliced by years, and please do add some report-wide insights.”
The report was built by Gen BI in just a few seconds. The report was visually appealing with a title page followed by insights statistics and analysis prompted by the user. The user can also change the style, for instance, by selecting ‘page insights’ text directly on the report which shows up a few options such as analytical tone.
4. Integrate Gen BI in host applications
Users can integrate Pyramid’s Gen BI into other applications without writing a line of code. Begin by building a dashboard on the Pyramid platform using the LLM-driven chatbot.
For example, when the chatbot asks the user to select a model, the user prompt can look something like this:
User: “Build me a quick slide, showing manufacturer sales by customer demographics. Do another chart showing me product subcategories in a scatter comparing returns versus quantities purchased. Lastly, show me sales by states on a map and slice it by month and promotion type.”
Gen BI (presents dashboard on screen): “The dashboard provides insights into manufacturer sales by customer demographics, a comparison of returns versus quantities a scatter plot, and sales by states in slicing options.”
Save the dashboard, at the bottom of the platform, select ‘my dashboard’ followed by selecting ‘extensions’ then select ‘embed code.’ Then inject the copied code straight into the host application.
Solution Overview: Gen BI
Gen BI is a user-friendly tool that enables quick actions to analyze data and create reports by conversing with the chatbot in natural language.
Pyramid Analytics launched Gen BI earlier this year enabling data teams, organizations, and business leaders with the quick ability to generate detailed and yet easily comprehensible insights by simply prompting the chatbot through speech or text. Users can interrogate and customize the data curated by Gen BI and also change the tone of the presentation with just a click of a finger.
About Pyramid Analytics
Pyramid Analytics is an award-winning Decision Intelligence Platform striving to equip people with AI-augmented, automated, and collaborative insights presented in a simplified manner for decision-makers.
With clients like Deloitte, Kellogs, and Hallmark, Pyramid Analytics enables a strategic, enterprise-wide approach to business intelligence and analytics. The firm’s paramount mission is building an end-to-end Decision Intelligence solution tailored for business users, data teams, and project managers.
