Skip to content

CoreGenerative AI Explained Writing Better Prompts RAG Deep Dive AI Evals: Building the Suite Evaluating LLM systems Jev (System One)pg-jevSecurity & governanceAI Security Prompt Injection (LLM01)AI Governance Agentic & Enterprise AI

FoundationsHow Agentic AI Works How Models Call ToolsProtocolsModel Context Protocol Agent-to-Agent (A2A)DesignLangChain and LangGraph Agent Memory Agent Planning Agent SkillsRuntimeAgent Harnesses Agent Sandboxes Inference ServingOperationsAgent Observability

AI Economics AI & Finance AI & Bloomberg

All Databases Database Comparison

Overview Installation

What is new REPACK Replication and WAIT Autovacuum Asynchronous I/O Monitoring Plan advice SQL and COPY Maintenance extensions

MariaDB MySQL SQL Server Oracle Database

MongoDB DynamoDB CouchDB

pgvector Milvus Qdrant

Time-series & analytical

TimescaleDB InfluxDB Snowflake

Key-value & cache

Memcached Valkey

All Data Engineering

Data pipelines Change data capture

Streaming platforms

Workflow orchestration

Storage & formats

Lakehouse and table formats

Data warehousing

Compute & transformation

Apache Spark dbt

Schema registry

Pooling & proxying

PgBouncer Pgpool-II

HA & load balancing

Patroni keepalived HAProxy for PostgreSQL

Coordination & observability

etcd Prometheus

All TechnologyAreasOpen Source Systems ArchitectureGuidesKubernetes LinuxReferenceDecision guides Concepts

All LearningFoundationsHow a model reads textAgent EngineeringHow agents work Agent harnesses Loop engineering Human in the loop Multi-agent systems Agent observabilityCurriculaDatabase Engineering AI Engineering Production Infrastructure Claude and Anthropic tooling Building Agents with LangChain and LangGraphGuidesGit & GitHub ADLC

Concepts Editorial standards Curated Resources

What Is New What changed Signals Technology radar Release timeline ResearchSiteBlog About

Core AI

All AI TopicsDeep dives across the AI landscape Generative AIWhat the model families are and how they generate output Prompt EngineeringStructuring instructions so a model does the intended task Context EngineeringWhat goes in the context window, and in what order RAGRetrieval-augmented generation, end to end AI EvalsBuilding the suite: cases, datasets, graders, RAG and agent evals Judge ReliabilityHow far an LLM judge can be trusted, with the measured bias figures

Agents & tool use

How Agentic AI WorksA visual guide to the model, tool and verification loop How Models Call ToolsThe request-and-response mechanism behind every agent action LangChain and LangGraphWhere a ready-made agent harness ends and a graph runtime begins Agent MemoryWorking, long-term, and episodic memory across a conversation Agent PlanningInterleaving reasoning with action instead of planning upfront Agent ObservabilityDiagnose failed runs and verify external outcomes

Protocols & runtime

Model Context ProtocolA shared way to connect a model to outside data and functions Agent-to-Agent (A2A)How independent agents discover, delegate and track work Agent SkillsPackaged procedures an agent loads only when needed Agent HarnessesThe layer that owns the agent loop, permissions and sandbox Agent SandboxesWhere unvetted model-generated code actually runs Inference ServingThe engine and the control plane behind a served model

Governance & risk

AI SecurityOWASP Top 10 for LLMs and the AI verification standard AI GovernanceNIST AI RMF, the EU AI Act, and ISO 42001 Agentic AIAgents, tool use, and the risks they introduce

Applied

AI EconomicsInference cost, utilisation, build versus buy AI in FinanceTrading, risk, RegTech, payments

Databases

DatabasesRelational, document, vector, graph, streaming Database comparisonPick up to three engines and compare them side by side PostgreSQLArchitecture, extensions, pgvector PostgreSQL InstallationUbuntu, Debian, RHEL family, Windows and macOS PostgreSQL 19REPACK, WAIT, logical sequences, plan advice, upgrade breakers

Data & infra

Data EngineeringPipelines, CDC, streaming, orchestration, lakehouse MiddlewarePgBouncer, Pgpool-II, keepalived, HAProxy Open SourceLinux, Kubernetes, cloud-native infrastructure SystemsOS, CPU, GPU, networking, distributed systems ArchitectureDesign patterns for AI and distributed systems

Reference

All TechnologyDatabases, middleware, systems, and infrastructure Technology mapHow every technology connects to the others ConceptsThe ideas the technology pages turn on Decision guidesWhat actually decides a technology choice

Foundations

How a model reads textOne sentence followed from raw text to a predicted next token: tokenisation, token IDs, embeddings, attention and sampling, and what each step decides about cost and behaviour.

Agent Engineering

How agents workFollow a document assistant from a user goal through model decisions, authorised tools, observations and verification, with a visual agent loop.Agent harnessesAn agent is a model plus a harness. What the harness is, the seven subsystems every one of them implements, and why the hardest parts are context, verification and the credential boundary.Loop engineeringDesigning the system that prompts the agent, rather than prompting it yourself: what a loop contains, what the first measurement of real adoption found, and why the evidence for the value claims does not yet exist.Human in the loopPausing an agent for human approval: how interruption and checkpointing actually work, what persistence does not guarantee, and the evidence that approval gates decay into rubber stamps.Multi-agent systemsWhen to split agent work, how adaptive routing differs from adding workers, and what published research says about coordination failures, verification and cost.Agent observabilitySeparate model proposals, tool attempts and real outcomes; follow an uncertain write through a redacted trace and turn the failure into a regression test.

Curricula

Database EngineeringA staged path from modelling a domain in tables to evolving a live schema under traffic, covering normalisation, storage internals, index design, the query planner, transactions and migrations, with a verifiable capability at each stage.AI EngineeringA staged path from using an LLM API correctly to shipping an evaluated, guard-railed RAG or agent system, with a verifiable capability per stage.Production InfrastructureA staged path from Linux fundamentals to running highly available, observable services on Kubernetes, with a verifiable capability at each stage.Claude and Anthropic toolingA staged path through the Messages API, tool use, Claude-specific evaluation, MCP, Claude Code permissions, and production cost, latency and failure.Building Agents with LangChain and LangGraphA runnable document-assistant course covering model interfaces, reducers, retrieval evaluation, bounded delegation and persistent approval, with an offline lab and an interactive state diagram.

Guides

Git & GitHubVersion control and collaboration, from commits to team workflows.ADLCDefine, design, build, evaluate, deploy, monitor and iterate.

Latest

What Is NewTechnical commentary on recent developments What changedCorrections, releases and what is worth watching

Signals & data

SignalsChanges in direction already visible in primary sources Technology radarWhere things sit, with the evidence for each position Release timelineReleases and end-of-life dates, generated from support data

Depth

ResearchWhere AI is headed and open research threads

Home
Technologies
PgCat

Middleware

PgCat

How PgCat connects to the rest of the stack: what it depends on, what depends on it, and what it is commonly deployed alongside.

Meta Intelligence Systems

Intelligence, not information: what a technology is, how it behaves, why it matters, and how it fails. Written from primary sources.

AI

All AI Topics
AI Security
AI Governance
Agentic & Enterprise AI
RAG

Technology

All Technology
Databases
Data Engineering
Middleware
Open Source
Systems

AI in Finance

AI Economics
AI & Finance
AI & Bloomberg

Learn

Learning Paths
Git & GitHub
ADLC

Intelligence

What Is New
What changed
Signals
Technology radar
Release timeline
Research
Architecture
Decision guides

Connect

Concepts
Discover
About
Editorial standards

© 2026 Meta Intelligence Systems

Search the site

Browse learning paths, technology or concepts while full-text search is unavailable.

MIS Assistant