AI platform Archives

Tech Optimizer

September 11, 2026

Improving Lakebase Postgres compute cache

Lakebase Postgres employs a disaggregated storage model that enhances data management through efficient caching, utilizing an object store like S3 for backups. The caching operates on two layers: within distributed storage for optimizing write and read performance, and on the Postgres compute side for ultra-fast access to frequently accessed pages. Traditional Postgres caching involves shared buffers and the OS page cache, which leads to double buffering and inefficiencies. Lakebase Postgres addresses these issues by implementing a local file cache (LFC) and larger shared buffers, allowing for more effective memory utilization without the drawbacks of the OS page cache. The shared buffers are set to a maximum of 1 GB, while the LFC can utilize up to 75% of DRAM. The introduction of huge pages reduces memory management overhead and improves performance, resulting in significant throughput increases and reduced latency in production environments. Recent enhancements have shown up to 2× throughput improvements and substantial reductions in CPU usage. The focus is now on extending these benefits to autoscaling Postgres computes, with plans to implement dynamic shared buffers and autoscaling huge pages.

AppWizard

September 5, 2026

Go grandmaster becomes first human to win series against advanced AI engine in 3-hour match

Go grandmaster Shin Jin-seo became the first human to win against the AI platform KataGo on July 21, achieving victory in an official three-game series with a final score of 2-1. Shin, who is 26 years old, played with a two-stone handicap and initially lost the first match before winning the next two. His victory is significant as it contrasts with the previous defeat of grandmaster Lee Sedol by Google's AlphaGo a decade ago. Shin expressed the desire to challenge AI under more difficult conditions in the future and emphasized the importance of personal strategy over imitating AI moves.

Tech Optimizer

September 4, 2026

Object Storage + WAL: Lakebase Postgres for the agentic era

Agents interacting with traditional OLTP databases face bottlenecks at the storage layer due to data movement required for operations like deploying instances and managing replicas. Object storage solutions, such as Amazon S3, provide a more efficient and cost-effective alternative. The text discusses two OLTP models: data-centric and transaction-centric. The transaction-centric model emphasizes the importance of transaction history, which Postgres retains through its write-ahead log (WAL). The WAL records every modification before it is committed, allowing for the reconstruction of the database state as of a specific point in time. Lakebase Postgres introduces a two-layer architecture separating compute and storage. The compute layer handles SQL operations, while the storage layer ensures data durability and historical integrity. The WAL is streamed to safekeepers for replication, and the pageserver transforms WAL into materialized pages stored in object storage. This architecture allows for efficient read and write operations without direct interaction with object storage during queries. Key features of Lakebase Postgres include branching, which allows agents to create pointers to specific LSNs without copying data; instant restore capabilities that enable point-in-time recovery; time travel queries for accessing historical data; and read replicas that do not require data duplication. The system also supports scaling down idle compute nodes and eliminates the need for separate copies of data for transactional and analytical processing by storing a single durable copy in object storage. Overall, Lakebase Postgres enhances agent interactions with the database by leveraging object storage as a foundational element.

Tech Optimizer

September 4, 2026

EDB Highlights Postgres AI Growth in Regulated Enterprise Markets

EnterpriseDB (EDB) has reported advancements in the adoption of its EDB Postgres AI (EDB PG AI) platform, which integrates transactional, analytical, and AI workloads. While 95% of enterprises aspire to become their own AI and data platforms within three years, only 13% have achieved this. EDB PG AI supports existing systems and prepares enterprises for future AI capabilities. A case study highlights PAC 2000A Conad, which restructured its data infrastructure with EDB PG AI, serving over 1,600 stores and 7,000 devices. In Korea, over billion has been invested in the sovereign AI market, with organizations seeking hybrid and on-premises solutions despite challenges from fragmented data environments. Notable users of EDB PG AI include the Industrial Bank of Korea and Shinhan EZ Insurance. EDB PG AI is built on Postgres and allows enterprises to optimize data and AI capabilities with governance at the data layer.

Tech Optimizer

September 1, 2026

Autoscaling Lakebase Postgres

Choosing a database instance size without prior knowledge of the workload can lead to inefficiencies and excessive compute usage. Lakebase Postgres addresses this with an autoscaling feature that eliminates manual sizing, utilizing in-place VM resizing and a monitoring algorithm for CPU, memory, and working set size. Lakebase Postgres separates compute and storage layers, allowing independent resizing of compute nodes without affecting the database. The autoscaling algorithm relies on three signals: CPU load (cpuGoalCU), memory use (memGoalCU), and compute-cache working set size (lfcGoalCU). The CPU load is monitored every five seconds, aiming to maintain it at or below 90% capacity. Memory usage is tracked at two frequencies: overall memory every five seconds and Postgres-specific memory every 100 milliseconds, with a goal to keep usage below 75% of allocated RAM. The compute cache evaluates active data access efficiency, adjusting size based on workload. The working set is estimated using a modified HyperLogLog algorithm that records timestamps for page accesses, allowing for distinct page estimates over various time frames. The algorithm projects future working-set growth to allocate sufficient cache while capping it at 75% of RAM. Resizing the compute involves four components: the autoscaler-agent, vm-monitor, Kubernetes scheduler, and NeonVM. Scaling up occurs when any of the three goals indicate a need for more resources, while scaling down includes verification to ensure sufficient memory remains for operations. Timely adjustments in both directions are prioritized to minimize costs.

Tech Optimizer

August 29, 2026

Overcoming the Barriers of Commercial DBs with Open Source and AI… EDB Hosts ‘PostgreSQL AI Summit’

A company's approach to artificial intelligence (AI) begins with determining the optimal structure for data storage and management, often facing challenges such as high costs of commercial databases and reliance on specific enterprise technologies. To effectively utilize AI, companies need to establish new infrastructures, including vector search capabilities, Search Augmentation Generative (RAG) systems, and data lakehouses. EDB is notable for connecting companies and technology partners that have transitioned to open-source databases. On September 3, EDB will host the 'EDB Postgres AI Summit Seoul 2026' at the Sofitel Ambassador Seoul, starting at 10:00 AM. This event is the largest PostgreSQL conference in Korea, themed 'Change the Game,' focusing on the shift from commercial databases to open-source and AI-driven frameworks, featuring real-world case studies. Approximately 300 C-level executives and IT decision-makers from various sectors are expected to attend, with participation by invitation only. The summit will include 15 sessions with customer case studies and technical presentations. Notable discussions will include IBK Industrial Bank of Korea's migration of 15 core systems to PostgreSQL, a semiconductor company's diversification of MPP databases on a DBaaS platform, and Kyobo Book Centre's database modernization strategy. Shopcast will present its development of an 'Agentic Lakehouse' integrating AI technology with a data lakehouse framework. The keynote address will be given by Kim Deok-joong, discussing organizational management strategies for integrating AI agents. The technical sessions will cover the architecture of the 'EDB Postgres AI' platform, which supports AI vector search and RAG, with live demonstrations of the analytics engine ClickHouse and LakeHouse technology for analyzing petabyte-scale data. EDB's domestic distributors and international partners will participate as sponsors. The summit aims to showcase technologies and case studies from the domestic ecosystem, addressing PostgreSQL adoption, system migration, operations, data analysis, and AI implementation. EDB manages transaction, analytics, and AI workloads using Postgres in cloud environments, serving over 1,500 global customers. Herve Timsit, EDB's Chief Revenue Officer, emphasized the event's focus on sharing tangible results and addressing the challenges of commercial databases while investing in AI infrastructure.

Tech Optimizer

July 30, 2026

EDB Says Postgres AI Tops Enterprise AI Data Platform Benchmarks

EnterpriseDB (EDB) has released benchmark results for its EDB Postgres AI (EDB PG AI), showing superior performance compared to various AI data platforms, including vector databases and managed Postgres platforms. The independent testing by McKnight Consulting Group indicated that EDB PG AI outperformed competitors in all evaluated metrics, including query speed, accuracy, and cost efficiency. Key findings include: - EDB PG AI achieved a median query latency of 50 milliseconds at 50 million vectors, making it the fastest platform tested, 80 times faster than Databricks and 21 times faster than MongoDB Atlas. - It demonstrated the highest accuracy for core vector search with a recall rate of 0.911, outperforming MongoDB by 26% and Databricks by 17%. - In a concurrent agent loop test, EDB PG AI completed tasks in 27 milliseconds, significantly faster than MongoDB Atlas (699 milliseconds) and Databricks (5.1 seconds). - EDB PG AI provided 76 times better price performance than Databricks and 34 times better than MongoDB. - The platform integrates vectors, structured data, and analytics within a single operational database, enhancing efficiency and reducing the risk of data inconsistency. The benchmark study assessed the performance of EDB PG AI against Amazon Aurora PostgreSQL, Databricks, Crunchy Bridge, MongoDB Atlas, and open source PostgreSQL, all standardized on equivalent enterprise hardware.

Tech Optimizer

July 2, 2026

EDB Named a Leader in Multimodel Data Platforms Evaluation

EDB has been recognized as a Leader in The Forrester Wave: Multimodel Data Platforms, Q2 2026, with EDB Postgres AI (EDB PG AI) achieving the highest scores in Vision, Innovation, Roadmap, and Partner Ecosystem criteria. EDB PG AI integrates transactional, analytical, and AI workloads into a unified platform, supporting open-source frameworks and enabling various deployment options. The platform features governance at the data layer and is designed for operational efficiency, allowing organizations to implement sovereign AI quickly. EDB PG AI can be deployed on-premises, in hybrid environments, or across cloud infrastructures, backed by partnerships with companies like Dell, IBM, and NVIDIA.

Tech Optimizer

July 2, 2026

EDB named Forrester Leader in multimodel data platforms

EDB has been recognized as a Leader in Forrester's Multimodel Data Platforms evaluation for Q2 2026 for its EDB Postgres AI platform, receiving the highest scores in Vision, Innovation, Roadmap, and Partner Ecosystem. The platform is designed to manage mixed translytical and AI workload demands, offering flexibility in deployment across on-premises, hybrid, and multi-cloud environments. EDB's recent product update introduced agentic database and converged analytics functionalities, reportedly accelerating database tuning by up to tenfold and reducing analytics ownership costs by as much as 58%. The platform is supported by a partner ecosystem that includes Dell, IBM, NVIDIA, Red Hat, and Supermicro, which plays a crucial role in influencing database purchasing decisions. EDB's roadmap focuses on advancements in GPU-accelerated workloads, semantic intelligence, governance, and knowledge graph functionalities. The emphasis on sovereign deployment aligns with organizations' needs for control over sensitive data amidst stricter regulations.

Tech Optimizer

June 26, 2026

EDB Postgres AI Launch: Shifting To Autonomous, Sovereign Data For The AI Era

EnterpriseDB (EDB) introduced the EDB Postgres AI (EDB PG AI) platform on June 23, 2026, designed for AI applications to operate directly on live data rather than outdated copies from cloud data lakes. The platform allows organizations to host AI models, live data, and enterprise regulations within their infrastructure, reducing vendor lock-in and protecting regulated data. The EDB PG AI platform features a self-optimizing system that transforms PostgreSQL into an autonomous database, monitoring over 200 metrics for automated tuning and scaling. EDB claims performance troubleshooting can be up to 10 times faster, with issues resolved in minutes instead of the traditional 60 to 90 minutes. It also includes a converged query interface that integrates various data types into a unified engine, enabling AI agents to access authorized live data. An agent governance framework will be introduced in late 2026 to address risks associated with AI operations. EDB collaborates with IBM Power for a robust AI-ready infrastructure and integrates Red Hat Ansible Automation Platform for enhanced management capabilities.