Glossary
Glossary
Section titled “Glossary”This glossary defines key terms and concepts used throughout the UniCraft documentation.
API Key
Section titled “API Key”A unique identifier used to authenticate API requests to UniCraft. API keys are used to identify your account and track usage.
Anthropic
Section titled “Anthropic”An AI company that provides Claude models, known for their helpfulness and safety features.
A/B Testing
Section titled “A/B Testing”A method of comparing two or more versions of a configuration to determine which performs better for a specific metric.
Batch Processing
Section titled “Batch Processing”Processing multiple requests together to improve efficiency and reduce costs.
Budget
Section titled “Budget”A spending limit set for AI usage over a specific time period (daily, weekly, monthly).
A temporary storage mechanism that stores frequently accessed data to improve performance and reduce costs.
Chat Completion
Section titled “Chat Completion”A type of AI request that generates conversational responses based on a series of messages.
Circuit Breaker
Section titled “Circuit Breaker”A design pattern that prevents cascading failures by automatically stopping requests to a failing service.
Claude
Section titled “Claude”AI models developed by Anthropic, including Claude 3 Haiku, Sonnet, and Opus.
Cohere
Section titled “Cohere”An AI company that provides Command and Embed models for various use cases.
Cost Optimization
Section titled “Cost Optimization”Strategies and techniques to minimize AI usage costs while maintaining quality.
Dashboard
Section titled “Dashboard”The web interface for managing UniCraft configuration, monitoring usage, and viewing analytics.
Data Protection
Section titled “Data Protection”Measures to secure and protect data, including encryption, access controls, and compliance.
Embedding
Section titled “Embedding”A vector representation of text that captures semantic meaning for use in similarity searches and other applications.
Error Rate
Section titled “Error Rate”The percentage of requests that fail or return errors.
Environment Variables
Section titled “Environment Variables”Configuration values stored in the system environment, used to configure UniCraft without hardcoding values.
Failover
Section titled “Failover”The automatic switching to a backup system when the primary system fails.
Fine-tuning
Section titled “Fine-tuning”The process of training a pre-trained model on specific data to improve performance for particular tasks.
Gemini
Section titled “Gemini”AI models developed by Google, including Gemini Pro and Gemini Pro Vision.
Generative Pre-trained Transformer models developed by OpenAI, including GPT-3.5 and GPT-4.
Health Check
Section titled “Health Check”A mechanism to monitor the status and availability of services and providers.
Hugging Face
Section titled “Hugging Face”A platform that provides access to various open-source AI models and tools.
Input Tokens
Section titled “Input Tokens”The number of tokens in the text you send to the AI model.
Integration
Section titled “Integration”The process of connecting UniCraft with other systems, tools, or platforms.
LangChain
Section titled “LangChain”A framework for building applications with Large Language Models (LLMs).
Load Balancing
Section titled “Load Balancing”Distributing requests across multiple providers or instances to improve performance and reliability.
An AI system trained to perform specific tasks, such as text generation, analysis, or embeddings.
Model Selection
Section titled “Model Selection”The process of choosing the most appropriate AI model for a specific task or use case.
A workflow automation tool that can be integrated with UniCraft to create AI-powered automations.
OpenAI
Section titled “OpenAI”An AI company that provides GPT models and other AI services.
Output Tokens
Section titled “Output Tokens”The number of tokens in the text generated by the AI model.
Pathways Language Model developed by Google.
Provider
Section titled “Provider”An AI service provider that offers models and APIs, such as OpenAI, Anthropic, or Google.
Prompt
Section titled “Prompt”The input text or instructions given to an AI model to generate a response.
Quality Score
Section titled “Quality Score”A metric that measures the quality of AI model responses.
Quality Threshold
Section titled “Quality Threshold”The minimum quality score required for a response to be considered acceptable.
Rate Limiting
Section titled “Rate Limiting”Controlling the number of requests that can be made within a specific time period.
Retry Logic
Section titled “Retry Logic”Automatically retrying failed requests to improve reliability.
Routing
Section titled “Routing”The process of directing requests to the most appropriate provider or model.
Routing Strategy
Section titled “Routing Strategy”The method used to select providers and models, such as cost-optimized, performance-optimized, or quality-optimized.
Smart Routing
Section titled “Smart Routing”Automatically selecting the best provider and model based on cost, performance, and quality criteria.
Starlight
Section titled “Starlight”The documentation framework used by UniCraft for building documentation sites.
Temperature
Section titled “Temperature”A parameter that controls the randomness of AI model responses (0.0 to 1.0).
A unit of text that AI models process, typically a word or part of a word.
A parameter that controls the diversity of AI model responses by limiting the probability mass.
UniCraft
Section titled “UniCraft”The AI model router platform that provides unified access to multiple AI providers.
Usage Analytics
Section titled “Usage Analytics”Data and insights about how AI models are being used, including costs, performance, and patterns.
Vector
Section titled “Vector”A mathematical representation of data in multi-dimensional space, used in embeddings and similarity searches.
Webhook
Section titled “Webhook”A mechanism for receiving real-time notifications when specific events occur in UniCraft.
Workflow
Section titled “Workflow”A series of automated steps that process data or perform tasks, often involving AI models.
Common Acronyms
Section titled “Common Acronyms”Application Programming Interface - a set of protocols and tools for building software applications.
Artificial Intelligence - computer systems that can perform tasks typically requiring human intelligence.
Large Language Model - AI models trained on large amounts of text data to understand and generate human-like text.
Machine Learning - a subset of AI that enables systems to learn and improve from experience.
Natural Language Processing - the field of AI focused on the interaction between computers and human language.
Representational State Transfer - an architectural style for designing web services.
Software Development Kit - a collection of tools and libraries for developing software applications.
User Interface - the visual and interactive elements of a software application.
User Experience - the overall experience of using a software application.
Technical Terms
Section titled “Technical Terms”Authentication
Section titled “Authentication”The process of verifying the identity of a user or system.
Authorization
Section titled “Authorization”The process of determining what actions a user or system is allowed to perform.
Encryption
Section titled “Encryption”The process of encoding data to protect it from unauthorized access.
Endpoint
Section titled “Endpoint”A specific URL or address where an API can be accessed.
JavaScript Object Notation - a lightweight data format used for API communication.
Hypertext Transfer Protocol - the protocol used for web communication.
HTTP Secure - HTTP with encryption for secure communication.
RESTful
Section titled “RESTful”Following REST principles for API design.
Webhook
Section titled “Webhook”A way for applications to receive real-time notifications about events.
Business Terms
Section titled “Business Terms”Cost per Request
Section titled “Cost per Request”The amount charged for each individual API request.
Cost per Token
Section titled “Cost per Token”The amount charged for each token processed by the AI model.
Monthly Active Users
Section titled “Monthly Active Users”The number of unique users who interact with the system in a month.
Request Volume
Section titled “Request Volume”The total number of API requests made over a specific time period.
Response Time
Section titled “Response Time”The time it takes for an API request to be processed and return a response.
Throughput
Section titled “Throughput”The rate at which requests are processed by the system.
Uptime
Section titled “Uptime”The percentage of time that a service is available and operational.
Security Terms
Section titled “Security Terms”API Key Rotation
Section titled “API Key Rotation”The practice of regularly changing API keys to maintain security.
Data Anonymization
Section titled “Data Anonymization”The process of removing or masking personally identifiable information from data.
Encryption at Rest
Section titled “Encryption at Rest”Encrypting data when it is stored on disk or in databases.
Encryption in Transit
Section titled “Encryption in Transit”Encrypting data when it is transmitted over networks.
IP Whitelisting
Section titled “IP Whitelisting”Restricting access to specific IP addresses or ranges.
Rate Limiting
Section titled “Rate Limiting”Controlling the number of requests that can be made within a specific time period.
Secret Management
Section titled “Secret Management”The practice of securely storing and managing sensitive information like API keys.
Monitoring Terms
Section titled “Monitoring Terms”A notification sent when a specific condition is met, such as high error rates or cost thresholds.
Dashboard
Section titled “Dashboard”A visual interface that displays key metrics and information about system performance.
Health Check
Section titled “Health Check”A mechanism to monitor the status and availability of services.
Logging
Section titled “Logging”The process of recording events and activities for debugging and monitoring purposes.
Metrics
Section titled “Metrics”Quantitative measurements of system performance and behavior.
Monitoring
Section titled “Monitoring”The continuous observation of system performance and health.
Service Level Agreement - a commitment to provide a certain level of service.
Service Level Objective - a target for system performance or availability.
Integration Terms
Section titled “Integration Terms”Custom Node
Section titled “Custom Node”A specialized component for workflow automation tools like N8N.
Integration
Section titled “Integration”The process of connecting UniCraft with other systems or platforms.
Software Development Kit - tools and libraries for integrating with UniCraft.
Webhook
Section titled “Webhook”A mechanism for receiving real-time notifications about events.
Workflow
Section titled “Workflow”A series of automated steps that process data or perform tasks.
Performance Terms
Section titled “Performance Terms”Caching
Section titled “Caching”Storing frequently accessed data in memory to improve performance.
Connection Pooling
Section titled “Connection Pooling”Reusing database or network connections to improve efficiency.
Load Balancing
Section titled “Load Balancing”Distributing requests across multiple servers or providers.
Optimization
Section titled “Optimization”Improving system performance through various techniques.
Performance
Section titled “Performance”The speed and efficiency of system operations.
Scalability
Section titled “Scalability”The ability of a system to handle increased load or demand.
Throughput
Section titled “Throughput”The rate at which requests are processed by the system.
Cost Management Terms
Section titled “Cost Management Terms”Budget
Section titled “Budget”A spending limit set for AI usage over a specific time period.
Cost Analysis
Section titled “Cost Analysis”The process of examining and understanding spending patterns.
Cost Optimization
Section titled “Cost Optimization”Strategies to minimize costs while maintaining quality.
Cost per Request
Section titled “Cost per Request”The amount charged for each individual API request.
Cost per Token
Section titled “Cost per Token”The amount charged for each token processed by the AI model.
Spending Limit
Section titled “Spending Limit”A maximum amount that can be spent on AI services.
Usage Analytics
Section titled “Usage Analytics”Data and insights about how AI models are being used.
Next Steps
Section titled “Next Steps”After familiarizing yourself with these terms:
- Explore Documentation: Use these definitions to better understand the documentation
- Ask Questions: Don’t hesitate to ask for clarification on any terms
- Stay Updated: Terms and concepts may evolve as the platform develops
- Share Knowledge: Help others understand these concepts
- Contribute: Suggest new terms or improvements to definitions