Tech & AI Glossary
61 terms across 12 categories
Agentic AI
AI systems that can autonomously plan, execute multi-step tasks, and interact with external tools and environments.
AI Agent ★
An autonomous software entity that perceives its environment, makes decisions, and takes actions to achieve specific goals.
AI Visual Process ★
The unified three-stage pipeline for visual AI: analyzing images (encoding pixels to semantic vectors), generating images (diffusion from noise to output), and iterating images (generative and deterministic editing).
Apache Kafka
A distributed event streaming platform for building real-time data pipelines and streaming applications at scale.
API (Application Programming Interface)
Set of rules and protocols that allow different software applications to communicate.
CDN (Content Delivery Network)
Distributed network of servers that delivers web content to users from geographically nearby locations.
CI/CD (Continuous Integration/Continuous Deployment)
Automated software development practice for frequent code integration and deployment.
Closure
A function that retains access to variables from its outer scope even after that scope has finished executing.
Cognitive Orchestration Engine ★
An architectural pattern for frontier AI systems that combines sparse mixture of experts, dynamic routing, tool delegation, and multi-tier memory to create general-purpose reasoning platforms rather than simple language models.
Context Window
The maximum amount of text (measured in tokens) that an LLM can process in a single interaction.
CORS (Cross-Origin Resource Sharing)
A security mechanism that allows or restricts web pages from making requests to domains different from the one serving the page.
Data Pipeline
Automated workflow that moves and transforms data from source systems to destination systems.
Decorator
A design pattern that allows behavior to be added to objects or functions dynamically without modifying their source code.
Design Pattern
A reusable solution template for commonly occurring problems in software design, providing proven approaches to structure code.
Docker ★
Docker is a revolutionary containerization platform that has transformed how developers build, ship, and run applications. By packaging software into standardized, isolated containers, Docker ensures consistent behavior across development, testing, and production environments.
Edge Computing
Computing paradigm that processes data near its source rather than in centralized data centers.
Embedding
Numerical vector representation of text, images, or other data for machine learning.
Embedding Server
A specialized server that converts text into vector representations (embeddings) for semantic search and RAG applications.
ETL (Extract, Transform, Load)
Data integration process that extracts data from sources, transforms it, and loads it into a destination system.
FastAPI ★
A modern, high-performance Python web framework for building APIs with automatic OpenAPI documentation.
Fine-tuning
Adapting a pre-trained model to specific tasks or domains by training on specialized data.
Gating Network
A neural network component that learns to route inputs to appropriate expert sub-networks in a Mixture of Experts architecture, determining which experts should process each token.
GGUF (GPT-Generated Unified Format)
File format for storing quantized LLM models, designed for efficient loading and inference with llama.cpp.
Git
Distributed version control system for tracking changes in source code during software development.
GraphQL
Query language for APIs that lets clients request exactly the data they need, developed by Facebook.
Hallucination
When AI models generate plausible-sounding but factually incorrect or fabricated information.
HTTPS/TLS
Secure communication protocol that encrypts data between browsers and servers using TLS certificates.
Hybrid Agent ★
An AI agent architecture that splits cognition between a cloud-based reasoning engine (the Architect) and local execution tools (the Builder), connected through a ReAct loop for iterative refinement.
Hybrid Retrieval ★
A search strategy that combines sparse retrieval (keyword matching from structured databases) with dense retrieval (semantic search from vector databases) to deliver both precise facts and contextually relevant results.
Inference Server
A server that hosts LLM models and processes requests to generate text responses, enabling local or self-hosted AI.
JWT (JSON Web Token)
Compact, URL-safe token format for securely transmitting information between parties as a JSON object.
Kubernetes ★
Open-source container orchestration platform for automating deployment, scaling, and management.
LLM (Large Language Model) ★
AI models trained on massive text datasets to understand and generate human-like text.
MCP (Model Context Protocol)
Anthropic's open protocol enabling AI models to securely connect with external data sources and tools.
Mixture of Experts ★
A neural network architecture that routes inputs to specialized sub-networks (experts), activating only a subset for each query to achieve massive scale with efficient computation.
Next.js
A React framework for building full-stack web applications with server-side rendering, static generation, and API routes.
Nginx
High-performance web server and reverse proxy known for its stability, speed, and low resource usage.
OAuth 2.0
Authorization framework enabling third-party applications to access user resources without sharing credentials.
ORM (Object-Relational Mapping)
Programming technique that converts data between object-oriented code and relational databases.
pgvector
PostgreSQL extension that adds vector similarity search capabilities for AI and machine learning applications.
PostgreSQL ★
Open-source relational database with advanced features like JSONB, full-text search, and extensions.
Prompt Engineering
The practice of designing and optimizing inputs to AI models to achieve desired outputs.
Prompt Processing Pipeline ★
The complete computational journey of a user prompt through an AI system: from text tokenization and embedding lookup, through transformer attention layers and expert routing, to final token prediction and response generation.
PWA (Progressive Web App)
Web application using modern APIs to deliver app-like experiences with offline support and installability.
Quantization
Technique to reduce LLM memory usage by representing model weights with lower precision numbers (e.g., 4-bit instead of 16-bit).
RAG (Retrieval-Augmented Generation) ★
AI technique combining vector search with LLMs to provide contextual answers from custom knowledge bases.
Rate Limiting
A technique to control the number of requests a client can make to an API or service within a specified time window.
REST API
Architectural style for designing networked applications using HTTP methods to perform CRUD operations on resources.
Serverless
Cloud execution model where the provider manages servers, scaling, and infrastructure automatically.
Sparse Activation
A computation strategy where only a subset of neural network parameters are activated for each input, enabling massive model scale while maintaining efficient inference.
SSH (Secure Shell)
Cryptographic network protocol for secure remote login, command execution, and file transfer.
SSR (Server-Side Rendering)
Rendering web pages on the server, sending fully-formed HTML to the browser for faster initial load.
SvelteKit ★
Full-stack web framework built on Svelte with server-side rendering and routing.
Technological Singularity
A hypothetical future point where technological growth becomes uncontrollable and irreversible, often associated with the emergence of superintelligent AI.
Token
The basic unit of text that LLMs process - typically a word, subword, or character.
Tokenization ★
The process of breaking text into smaller units called tokens (words, subwords, or characters) that AI models can process numerically, using algorithms like Byte Pair Encoding (BPE) or SentencePiece.
Transformer ★
Neural network architecture using self-attention mechanisms, the foundation of modern LLMs like GPT and Claude.
TypeScript
Strongly-typed superset of JavaScript that compiles to plain JavaScript, adding static type checking.
Vector Database
Specialized database for storing and searching high-dimensional vector embeddings.
VPS (Virtual Private Server)
Virtualized server instance providing dedicated resources on shared physical hardware.
WebSocket
Protocol providing full-duplex, bidirectional communication channels over a single TCP connection.