Fujitsu remains committed to exploring cutting-edge AI technologies while continuously strengthening Fujitsu Enterprise Postgres as a robust and reliable database management system. With the release of version 17 SP1, we introduce enhancements in Knowledge Data Management, alongside improvements in database monitoring and job scheduling. These updates streamline data operations, optimize AI workflows, and enhance system efficiency
Key innovations in Fujitsu Enterprise Postgres 17 SP1
Knowledge Data Management
Fujitsu Enterprise Postgres' approach to Generative AI
In recent years, the use of generative AI tailored to specific industries and fields has been increasing, driven by companies leveraging their own data. Retrieval-Augmented Generation (RAG) is gaining attention as a technology that enables this. RAG improves the accuracy of responses by searching external data sources for information relevant to the question and then passing both the question and the relevant information to Large Language Models (LLMs).
Fujitsu Enterprise Postgres is committed to enhancing the functionality required of external data sources for RAG, providing a data infrastructure that is reliable, easy to use, and RAG-compatible.
A set of capabilities to meet the AI revolution
Fujitsu Enterprise Postgres has a long history in enhancing database security, performance, and reliability with a set of unique features that enhance PostgreSQL. And with its latest release, Fujitsu Enterprise Postgres offers Knowledge Data Management to support the rapidly evolving field of generative AI.
In RAG, knowledge data can be represented in text, graph, or vector format, for high-speed search. Knowledge Data Management supports the utilization of knowledge data by streamlining semantic relationship-based search using vectors and graphs, knowledge data management, and RAG application development and operation. You can now search knowledge data in three ways –text semantic search, vector similarity search, and graph traversal–, in addition to traditional search methods.
For details on the Knowledge Data Management, check our online manual Knowledge Data Management Feature User's Guide.Centralized data management - Text, vector, and graph data in the same database
Vector data and graph data have different formats from traditional structured data, so they are typically managed in dedicated databases. Apart from traditional data formats, Fujitsu Enterprise Postgres can manage knowledge data in both formats, eliminating the need for dedicated databases for each format.
Knowledge Data Management handing of various data formats
Moreover, knowledge data can be securely managed using Fujitsu Enterprise Postgres' existing security features, such as Transparent Data Encryption, Data Masking, Dedicated Audit Log, FIPS compliance, confidentiality management, and policy-based login security.
For information on integrating with existing security features, check our blog post Protecting organizational data with Fujitsu's security technology.Text semantic search
Semantic search typically involves converting both the search data and the query into vectors. The search data is vectorized and stored in a vector database. When a query is made, the RAG application vectorizes the query text, which is then used to find similar vectors within the database. Consequently, any updates to the original text data necessitate corresponding updates to the stored vectors. Fujitsu Enterprise Postgres streamlines this process with Knowledge Data Management by converting queries to vectors and calculating similarity, thus enabling efficient retrieval of highly relevant information.
Furthermore, data is automatically vectorized to keep it up to date. This makes it possible to perform semantic searches simply by entering text as a query. In addition, there is no longer any need to perform vectorization on the RAG application side or to update vector data when updating data, reducing the cost of developing RAG applications and updating vector data.
Knowledge Data Management handling of text semantic search
RAG application development support
Fujitsu Enterprise Postgres can be utilized as a knowledge base for RAG because it handles vector data (through the pgvector extension) and graph data (via Apache AGE). To allow AI application development, you can utilize Python frameworks such as LangChain.
For more information on developing AI applications utilizing Python, check our brochure Knowledge Data Management utilization with LangChain and Python.
Database monitoring using Datasentinel
Datasentinel is a comprehensive monitoring platform that provides efficient visualization and advanced monitoring of your database environment. Datasentinel servers and databases work together to provide centralized information about database health and performance, enabling accurate analysis of cluster activity, automatic notification of problems, and rapid identification of resource-intensive processes.
The platform helps provide real-time insights on database performance, and provides streamlined and comprehensive visualization with a wide range of features, such as session history, table & index metrics, lock explorer, top queries, real time view, and server, instance, and database metrics.
This empowers database administrators and IT professionals by ensuring optimal performance and availability of Fujitsu Enterprise Postgres instances.
Datasentinel monitoring of Fujitsu Enterprise Postgres activity
To learn how to monitor Fujitsu Enterprise Postgres using Datasentinel, check the platform's official website.SQL Job Scheduler
Fujitsu Enterprise Postgres now supports the SQL job scheduler pg_cron, so you can schedule SQL jobs within the database, allowing application developers to periodically run jobs that can be executed by SQL or stored procedures for routine maintenance, such as VACUUM.
For details on how to use pg_cron with Fujitsu Enterprise Postgres, check our online documentation Installation and Setup Guide for Server > pg_cron.Python client
Leveraging Python simplifies the process of embedding AI into applications, given the language's strong support for AI-related tasks. And now Fujitsu Enterprise Postgres supports psycopg3 as a Python client.
You can check how to access Fujitsu Enterprise Postgres via Python in our online page Application Development Guide > Python Language Package (psycopg).Final thoughts
With version 17 SP1, Fujitsu Enterprise Postgres delivers powerful enhancements that simplify AI-driven knowledge management while improving fundamental, enterprise functionality in database monitoring and job execution. These updates reinforce Fujitsu Enterprise Postgres as a leading solution for organizations looking to optimize their data infrastructure for AI applications and efficient database operations.
Try Fujitsu Enterprise Postgres 17 SP1 today
Experience the latest advancements in AI-powered knowledge data management and robust database operations by trialing a full-featured version of Fujitsu Enterprise Postgres 17 SP1 for 90 days. Explore how these enhancements can optimize your AI applications and database usability.
Don't forget to subscribe to the blog, and we will keep you informed when a new post goes live.
is an enhanced distribution of PostgreSQL, 100% compatible and with extended features.
