Evolution of Business Intelligence and Intelligent BI: Overview of Sugar BI’s DI Predictive Module

This article, originally shared at the DataFun Enhanced Analytics Forum in December 2022, introduces the development history of Business Intelligence (BI), the emergence of intelligent BI, and the capabilities of Sugar BI’s DI (Intelligent Prediction) module.

Key points of the presentation:

1. Introduction to the evolution of BI, detailing each stage of change. 2. Discussion of trends in the intelligent BI era, Sugar BI’s strengths in intelligent BI, and a comparison of various predictive‑analysis platforms. 3. Demonstration of Sugar BI’s intelligent prediction DI module and its application scenarios.

BI Development Timeline

• 1865 – The term “business intelligence” first appeared in a book by banker Henry, emphasizing data collection and analysis. • 1958 – IBM scientist Hans (often called the “father of BI”) described the value and potential of BI in a business intelligence system. • 1989 – Gartner formally defined BI as encompassing data storage and analysis. • Since the 1990s, BI has progressed through three phases: BI 1.0, BI 2.0 and BI 2.5.

BI 1.0

BI 1.0 emerged with data warehouses. Business data were centralized, and users accessed data via offline methods (e.g., static Excel reports). The process required developers to export data from databases, then analysts manually created charts, leading to significant latency and high labor costs. Building a visual page often took about a month, and any metric change required another week of work.

BI 2.0

With the rise of internet technologies and IM tools, data became more timely and complex, enabling online collaboration and self‑service analytics. Users could create interactive dashboards within minutes, but the quality of analysis varied due to differing skill levels, and decision‑making still faced limitations.

We are now in a transitional “BI 2.5” stage, where agile BI is mature but intelligent BI is still evolving.

Intelligent BI Era

AI is reshaping BI, adding a layer of intelligence through enhanced analytics (Gartner, 2017). This includes automated data preparation, insight discovery, and sharing, allowing non‑technical users to leverage machine learning without writing code.

Key intelligent features in Sugar BI:

Intelligent chart recommendation – the system suggests the most suitable chart type for a given dataset.

Automatic analysis – one‑click generation of interactive reports using over a hundred built‑in chart‑data matches.

Intelligent interaction – voice‑enabled query and control across mobile, PC, and large‑screen devices.

Intelligent decision – machine‑learning‑based predictive analysis to support leadership decisions.

Predictive Analysis Platforms

Three main categories:

BI platforms with built‑in prediction (e.g., Sugar BI’s DI module) – low‑code, suitable for business users.

Open‑source machine‑learning tools – powerful but require higher expertise.

Integrated AI development platforms (e.g., Baidu BML) – API‑driven, aimed at AI experts.

Sugar BI’s DI Module

The DI module targets business users, requiring no coding. It integrates prediction results directly into visualizations, supporting structured data only. Users can create prediction fields through a four‑step workflow: select model → set parameters → configure input/output → generate field.

DI offers three model types:

Built‑in models – clustering and linear regression, automatically selected for optimal performance.

Training models – users define training and validation sets, choose algorithms, and publish the best model.

Custom/external models – upload or connect third‑party prediction services, with future Open‑API support.

Demo screenshots illustrate the built‑in model creation, training workflow (including binary classification), and the resulting visual dashboards.

Q&A Highlights

Q1: How does AI Q&A work without training data? A: The system uses built‑in NLP capabilities to map spoken words to database fields, enabling instant chart generation and allowing users to add synonyms for better accuracy.

Q2: Where can users learn more about the product? A: Detailed tutorials and documentation are available on the Sugar BI website, with a one‑month free trial.

Overall, the presentation provides a comprehensive overview of BI’s evolution, the shift toward intelligent, AI‑augmented analytics, and practical guidance on leveraging Sugar BI’s predictive DI module.