semantic search Archives

Tech Optimizer

July 12, 2026

Running pgvector in production on Amazon Aurora PostgreSQL

Running pgvector on Amazon Aurora PostgreSQL-Compatible Edition offers a vector store with operational capabilities, high availability, and scalability. It is favored for Retrieval Augmented Generation (RAG) workloads transitioning to production, but increased traffic introduces challenges like query latency and memory management. Key operational practices for pgvector workloads include selecting the appropriate index type (HNSW or IVFFlat), establishing a baseline schema, choosing a suitable distance operator, scaling the index through quantization and partitioning, and preparing for churn and observability. The prerequisites for using pgvector include an Aurora PostgreSQL-Compatible cluster with specific PostgreSQL versions and the vector extension enabled. The embedding model used in examples is Amazon Titan Text Embeddings V2, which produces 1024-dimensional embeddings. pgvector supports two Approximate Nearest Neighbor (ANN) index types: HNSW, which is efficient for querying and allows for incremental insertions, and IVFFlat, which is less resource-intensive but requires rebuilding if data changes. There are scenarios where forgoing an index is beneficial, such as small datasets or partitioned datasets requiring 100% recall. A baseline schema for a multi-tenant document store includes creating a table for documents with an embedding vector and establishing indexes for tenant IDs and embeddings using HNSW. The recommended parameters for HNSW include m = 16 and ef_construction = 128. Scaling to millions of vectors involves quantization, tuning HNSW parameters, and partitioning. Aurora Optimized Reads can extend effective cache capacity, and managing index churn is crucial for maintaining performance. Observability metrics include query-level statistics, instance-level metrics, and custom application-defined metrics. To clean up after testing, it is advisable to drop the created indexes and tables, and delete the Aurora PostgreSQL-Compatible cluster and any manual snapshots taken during testing.

Tech Optimizer

June 6, 2026

PostgreSQL vs SQL Server 2026: $0 vs $15K License

PostgreSQL has become the most utilized database for developers for two consecutive years as of 2026, while SQL Server has introduced native AI capabilities and vector search functionalities in its latest release. PostgreSQL is favored for new projects due to its permissive licensing, cross-platform operation, and extensive ecosystem of extensions, whereas SQL Server is preferred by organizations already using Microsoft products for its robust tooling and enterprise support. As of June 2026, SQL Server ranks third in the DB-Engines popularity ranking with a score of 698, while PostgreSQL ranks fourth with a score of 688. Developer adoption trends show that 49% of respondents in the Stack Overflow 2024 Developer Survey use PostgreSQL, compared to 27% for SQL Server. The latest stable version of PostgreSQL is 18.4, released on May 14, 2026, while SQL Server 2025 (version 17.x) became generally available on November 18, 2025. PostgreSQL is free under a permissive license, while SQL Server's Enterprise Edition can cost upwards of ,123 per two-core pack, with an eight-core minimum. PostgreSQL supports various operating systems, including Linux, Windows, macOS, and BSD, while SQL Server operates on Windows and Linux. PostgreSQL uses PL/pgSQL as its primary procedural language, allowing for multiple procedural languages, whereas SQL Server uses T-SQL. PostgreSQL's JSONB type offers rich indexing, while SQL Server added native JSON support in its 2025 release. Both databases are capable of handling high transaction volumes, but PostgreSQL excels in read-heavy scenarios due to its multi-version concurrency control, while SQL Server's columnstore indexes provide advantages for analytical workloads. PostgreSQL supports tables up to 32 TB, while SQL Server can handle a maximum database size of 524 PB. SQL Server's high availability technology includes Always On Availability Groups, while PostgreSQL offers built-in streaming and logical replication. The security features of both databases meet enterprise standards, with PostgreSQL providing a transparent patching process and SQL Server integrating with Microsoft tools. Real-world usage examples include Stack Overflow using SQL Server, while Instagram and Reddit utilize PostgreSQL. Developer sentiment increasingly favors PostgreSQL for new projects, emphasizing flexibility and open-source solutions. For new startups or greenfield projects, PostgreSQL is recommended due to its zero license cost and compatibility. Existing Microsoft enterprises are advised to stick with SQL Server for its integration benefits. Migration from SQL Server to PostgreSQL is becoming common, with a structured approach required for successful transitions.

Winsage

April 28, 2026

I investigated Windows 11’s massive 5GB monthly .msu updates. AI is only part of the problem.

Windows 11 updates have significantly increased in size, with monthly cumulative updates often exceeding 4GB and some approaching 5GB. One update can expand to nearly 9GB when extracted. Microsoft has shifted to delivering Latest Cumulative Updates (LCUs), which include all previous fixes, leading to larger update sizes over time. The introduction of Checkpoint Cumulative Updates aims to reduce this growth by establishing periodic baselines, but the effectiveness has been mixed. The May 2025 cumulative update saw a size increase from approximately 6.5GB to nearly 9GB, with new MSIX files related to semantic search and on-device AI contributing to this growth. Windows Update uses applicability logic to minimize download sizes for users, but enterprises must download full packages, resulting in increased storage costs. The average yearly storage cost for enterprises rose from about 11 GB in 2024 to 52 GB by 2026. Users can check their actual download sizes through the Windows Update settings and Event Viewer logs.

Tech Optimizer

April 28, 2026

Constructive Open Sources Agentic DB, the Postgres Memory Layer for AI Agents

Constructive has released agentic-db, an open-source Postgres database aimed at improving AI agents with features like persistent memory, structured knowledge, and hybrid retrieval. This database allows for efficient searching, filtering, and ranking of information, addressing inefficiencies associated with traditional markdown file storage. Key features of agentic-db include long-term memory, conversation tracking, a versioned registry for skills and tools, rules and policies for governance, task orchestration, and runtime observability. It is delivered as a single installable Postgres schema, indexed for multiple retrieval modes, and supports integration with various AI assistants through generated Agent Skills and CLIs. agentic-db is available under the MIT license for local use, with a cloud offering in development for secure, scalable solutions. Developers can access it on npm and GitHub.

Tech Optimizer

April 1, 2026

EDB Postgres AI for WarehousePG: Reclaiming control of the enterprise data warehouse

Many enterprises are transitioning from traditional proprietary data warehouse platforms, such as Teradata and Snowflake, due to issues like vendor lock-in, unpredictable pricing, and limited flexibility. EDB Postgres® AI (EDB PG AI) offers WarehousePG, an open-source, petabyte-scale data warehouse built on Postgres, designed to provide control, predictability, and data sovereignty while maintaining performance. WarehousePG utilizes a massively parallel processing (MPP) architecture, allowing it to efficiently execute complex queries across large datasets. It offers predictable performance without proprietary constraints, enabling organizations to avoid vendor-controlled execution engines. WarehousePG supports hybrid storage and SQL access to external data lakes through the Platform Extension Framework (PXF), simplifying ETL processes. It includes FlowServer for real-time data ingestion and supports in-database AI and machine learning, allowing for advanced analytics without transferring data outside the warehouse. The platform is engineered for high availability and enterprise readiness, with features like workload management and observability. Migration from legacy platforms is facilitated through a low-risk modernization pathway. Overall, WarehousePG provides a modern alternative to traditional data warehouses, emphasizing architectural control and open-source economics.

Winsage

March 24, 2026

A closer look at the Windows 11 features businesses rely on most

Windows 11 is a foundational platform for businesses, offering integrated productivity tools and compatibility with Microsoft 365, which enhances cybersecurity and IT management. Key security features include the requirement for TPM 2.0, Windows Hello for Business for secure sign-ins, and BitLocker device encryption enabled by default. The integration of Microsoft 365 Copilot allows AI-driven assistance for document creation and data analysis, streamlining workflows across applications like Word, Excel, Outlook, and Teams. IT management is simplified with tools like Windows Autopilot for zero-touch provisioning and Microsoft Intune for remote device management. Windows 11 also improves performance with features like adaptive energy saver mode and Fast Startup, while plans to enhance user experience include reintroducing the ability to move and resize the Taskbar.

Winsage

March 16, 2026

Microsoft Cancels Several Planned Copilot Integrations in Windows 11

Microsoft has halted plans to integrate Copilot into Windows 11 system interfaces, including notifications, Settings, and File Explorer. These features, initially announced in 2024, never reached the market. A Microsoft spokesperson indicated that feature development may involve private previews and public iterations, with potential changes or removals based on customer feedback. The anticipated Copilot features did not materialize, and after the delay of Windows Recall, resources were reallocated to address related issues. AI functionality later appeared in Settings and File Explorer without the Copilot branding, with new features such as semantic search and an AI actions menu that delegate tasks to other applications. The feature for Copilot suggestions in notifications has been canceled, and Microsoft has no plans to release it under the Copilot name. The company is shifting away from the Copilot branding in Windows 11, focusing on Microsoft 365 integrations and minimizing AI presence in response to user feedback and stock value concerns. Future AI features will be optional and easier to disable.

Winsage

March 11, 2026

Windows 12 could be the tipping point that finally pushes you to Linux

The anticipated release of Windows 12, rumored for 2026, has sparked speculation about a potential migration of users to Linux. Microsoft’s history with Windows shows a cycle of launching resource-intensive operating systems, leading to user dissatisfaction and exploration of alternatives. The end of support for Windows 10 has increased interest in open-source systems like Linux. The integration of artificial intelligence in Windows 12 may complicate user experiences, potentially driving users away due to increased demands and complexities. In contrast, Linux is evolving to offer a user-friendly experience without advertisements or the need for costly hardware upgrades, making it an appealing alternative for disenchanted Windows users.

Tech Optimizer

December 20, 2025

Watch Out Elasticsearch! Tiger Data’s PostgreSQL BM25 Search Extension Goes Open Source

Tiger Data has released pg_textsearch as an open-source extension under The PostgreSQL License on GitHub, previously exclusive to Tiger Cloud. This extension allows developers to perform BM25 relevance-ranked keyword searches directly within PostgreSQL, eliminating the need for external search systems like Elasticsearch. Notable features of pg_textsearch include support for 29+ languages, compatibility with partitioned tables, intuitive SQL syntax, a memtable architecture for optimized indexing, and adjustable BM25 ranking parameters. Tiger Data aims to develop a comprehensive Postgres Search Stack that integrates keyword and vector search capabilities.

Winsage

December 19, 2025

Upgrade to Microsoft Windows 11 Pro at 94% off

Microsoft Windows 11 Pro is available for purchase at a discounted price of .97, down from the regular price of 9, representing a 94% discount. The operating system offers features such as robust security, professional productivity tools, exceptional gaming capabilities, and advanced AI features through Microsoft Copilot.