Types of Temporal Data - Feature Engineering

Feature Factory: A Paradigm Shift for Enterprise Data

May 2, 2023

The world of enterprise data applications such as Business Intelligence (BI), Machine Learning (ML), and Artificial Intelligence (AI) is becoming increasingly critical for organizations of all sizes. As these technologies advance, businesses face the challenge of choosing the best tools to apply in different situations. The landscape of BI, ML, and AI tools has become commoditized and fragmented, leading to the development of various tools for specific purposes. Despite this, data application development in enterprises often remains siloed, and the full potential of data is not extracted. Enter Feature Factory – a new paradigm that aims to revolutionize open and scalable enterprise BI, ML, and AI development.

A Commoditized and Fragmented Landscape

The Enterprise ML and AI tools market, as well as the BI tools market, is highly commoditized, with a wide variety of tools available for use, each with its strengths and weaknesses. The tools range from open-source libraries for general ML purposes to platforms integrated with big cloud providers such as AWS SageMaker, Azure ML, and Google Vertex AI to tools embedded into specific business platforms such as Salesforce Einstein and SAP Predictive Analysis. The result is a fragmented market, with many different tools available, each with unique features and capabilities.

In this scenario, “which tool is better” is no longer relevant. Instead, organizations must focus on using the right ML tools for the right situations. However, the wide range of options available has not eliminated the issue of siloed development within enterprises.

Enterprise Data Solution Development: Siloed and Underutilized

Despite the growth in the use of BI, ML, and AI, there are still many challenges facing organizations when it comes to the development of these systems. One of the biggest challenges is the siloed nature of BI, ML, and AI development.

Organizations typically collect and store data from various sources, including ERP, CRM, digital marketing, and IoT. However, this data is rarely combined and, instead, is curated for each specific business domain. This means that the curated data, or feature tables, critical for ML/AI development, are prepared separately from each data source.

As a result, BI, ML, and AI development is very siloed, with different teams working on different data sources, using various tools, and with little or no collaboration between teams. This leads to a situation where the full power of the data is not being extracted, and organizations are missing out on the benefits that can be achieved through a more integrated and collaborative approach to ML and AI development.

Siloed Feature Discovery & Creation

Introducing Feature Factory: A New Paradigm for Open and Scalable Enterprise Data Solutions

Feature Factory is a groundbreaking concept that aims to address the issue of siloed development and unlock the full potential of enterprise data solutions. It is a data-driven mechanism designed to generate and manage feature spaces, queries, and tables.

dotData Feature Factory, Centralized Feature Discovery

Organizations can access a 100 times larger feature space by implementing Feature Factory. This enables data scientists to quickly understand the data tables and columns containing relevant signals for their problems and experiment with many more feature hypotheses and ideas. As a result, the quality of BI, ML, and AI solutions is significantly improved, and the efficiency of the development process is increased.

One of the critical advantages of Feature Factory is its ability to transform features from one-time ad-hoc queries into accumulated and reusable assets. It achieves this by providing a standard feature description language that eliminates the need for writing complex SQL queries. In addition, Feature Factory memorizes all the steps from raw data to feature tables so that anyone can reproduce them easily.

The Benefits of Feature Factory

  • Enhanced Collaboration: Feature Factory promotes collaboration between data scientists, engineers, and domain experts by providing a standardized approach to feature discovery.
  • Increased Efficiency: By automating the feature generation process and reducing the need for complex SQL queries, Feature Factory accelerates the development of BI, ML, and AI solutions.
  • Reusability and Reproducibility: Feature Factory enables organizations to accumulate and reuse features across different projects, saving time and resources. Moreover, its ability to memorize the entire feature generation process ensures reproducibility and knowledge retention.
  • Data Integration: Feature Factory breaks down data silos by allowing data from different sources to be combined and used effectively in BI, ML, and AI applications. This results in a more comprehensive understanding of the available data and higher-quality insights.
  • Scalability: Feature Factory is designed to be easily scalable, allowing organizations to adapt and grow their data solution capabilities as needed.
  • Transparency and Interpretability: The standard feature description language and memorization of feature generation steps make it easier for stakeholders to understand and interpret the BI, ML, and AI solutions, fostering trust in the technology.


The landscape of BI, ML, and AI tools is undoubtedly fragmented and commoditized, with many solutions available for a wide range of applications. However, the problem of siloed development within enterprises prevents organizations from leveraging the full power of their data.

Feature Factory is set to revolutionize the enterprise data landscape by offering a new paradigm for open and scalable data solution development. By promoting collaboration, increasing efficiency, improving solution quality, and breaking down data silos, Feature Factory enables organizations to unlock the true potential of their data.

As data-driven technologies continue to transform industries, adopting solutions like Feature Factory will be crucial for organizations that want to stay ahead of the curve and make the most of the opportunities these technologies present. By embracing Feature Factory, businesses can enhance their BI, ML, and AI capabilities and drive innovation and growth.

Learn more about how your organization could benefit from the powerful features of dotData by signing up for a demo.

Share On

Ryohei Fujimaki, PhD.

Ryohei Fujimaki

Ryohei is the Founder & CEO of dotData. Prior to founding dotData, he was the youngest research fellow ever in NEC Corporation’s 119-year history, the title was honored for only six individuals among 1000+ researchers. During his tenure at NEC, Ryohei was heavily involved in developing many cutting-edge data science solutions with NEC’s global business clients, and was instrumental in the successful delivery of several high-profile analytical solutions that are now widely used in industry. Ryohei received his Ph.D. degree from the University of Tokyo in the field of machine learning and artificial intelligence.