17.1 Glossary of Terms

This glossary serves as a reference for the technical, platform specific, and industry related terms used throughout the Craft whitepaper. It ensures clarity and consistent understanding for readers across all professional backgrounds.

• AI Model: A machine learning algorithm trained on large datasets to perform tasks such as text generation, image creation, or audio synthesis.

• Prompt Engineering: The process of designing inputs (prompts) that effectively guide AI models to produce desired outputs.

• Modular Tools: Individual functional units that can be dynamically loaded and used within Craft's ecosystem.

• Persistent Memory: A feature that enables the AI to remember user preferences, previous prompts, and outputs across sessions.

• Prompt Fatigue: The degradation of quality or consistency in AI responses over time due to repeated or poorly constructed prompt inputs.

• Fine-Tuned Model: A base AI model that has been trained further on a specialized dataset for specific performance improvements.

• API-First Architecture: A software design strategy that prioritizes building an application’s backend using APIs, allowing greater scalability and integration.

• Artificial Intelligence (AI): A branch of computer science focused on building systems capable of performing tasks that typically require human intelligence. These include natural language understanding, decision-making, computer vision, and more. Craft leverages AI to automate and enhance creative processes.

• Large Language Models (LLMs): Advanced neural networks trained on massive datasets to understand and generate human-like language. Examples include OpenAI’s GPT-4, Anthropic’s Claude, and Meta’s LLaMA. Craft integrates LLMs to power writing, summarization, chatbot interactions, and content generation.

• Prompt Engineering: The practice of designing and structuring inputs (prompts) to AI models in a way that guides them toward producing the most accurate, relevant, and context-aware outputs. Craft uses proprietary prompt engineering techniques to ensure optimal results across tools.

• Prompt Memory: Craft’s proprietary feature that enables intelligent context retention across sessions. It allows AI agents and tools to remember user preferences, tone, brand voice, and past interactions to deliver more personalized and coherent outputs.

• Modular Tools: Self-contained applications or functionalities within the Craft ecosystem that can be individually loaded, updated, or connected in workflows. Each tool serves a specific purpose—such as writing, video editing, or voice dubbing—and can be used independently or in combination.

• Workflow Automation: A visual and logic-driven system that enables users to connect multiple AI tools in a sequence. For example, a user can generate a blog, convert it into a voiceover, and compile it into a video—all without manual switching between platforms.

• Agent: A customizable AI persona within Craft that can be assigned a specific role, memory, personality, and task list. Agents are capable of completing complex multi-step tasks with contextual awareness, memory recall, and adaptive behavior.

• Text-to-Speech (TTS): A synthesis technology that converts written text into spoken audio. Craft’s TTS engine supports multilingual output, character-specific tones, and even cloned voices for consistent branding or personalization.

• Speech-to-Text (STT): Technology that transcribes spoken audio into written text. Used in tools such as Craft’s meeting summarizer and voice input tools to generate documents, transcripts, or notes automatically.

• Image Generation: AI-driven visual synthesis that translates text prompts into images. Powered by models like Stability AI and OpenAI’s DALL·E, Craft’s image tools enable creators to generate visuals for concept design, storyboards, and content production.

• Video Generation: The process of transforming scripts, text, or storyboards into video content using AI. Craft allows users to build videos from text, combine voiceovers, and animate scenes with minimal manual input.

• Multimodal AI: A category of artificial intelligence models capable of understanding and generating content across multiple formats—such as text, images, audio, and video. Craft’s ecosystem integrates multimodal capabilities to enable cross-format workflows.

• Persistent Memory: An intelligent system that retains context across sessions, enabling agents and tools to recall user-specific data such as tone of voice, frequently used topics, project history, and preferences.

• Fine-Tuned Models: Models that have been adapted using domain-specific or task-specific data to optimize performance in a certain area. Craft supports integration of fine-tuned LLMs for advanced use cases like legal writing, marketing copy, or educational content.

• Frontend: The part of the Craft platform that users interact with directly, built using modern technologies like React and Tailwind. It includes dashboards, editors, and all user interface components.

• Backend: The infrastructure that powers the core logic, APIs, and integrations of the Craft platform. Built with Node.js, Python, and containerized microservices, it manages data processing, tool orchestration, and model execution.

• API (Application Programming Interface): A standardized method for different software components to communicate. Craft’s platform offers internal and external APIs to power modular tools, integrate external services, and enable metered usage models.

• Template Engine: A framework within Craft that allows users to generate content based on pre-defined prompts and structures. Templates can be reused, customized, and scaled to automate content creation.

• Batch Processing: A mode in Craft that allows users to run AI tools on multiple pieces of content at once. Ideal for agencies or enterprises that need to process large volumes of data or generate bulk outputs efficiently.

• Brand Voice: A user-defined tone, style, and personality that is consistently applied across all content generated by Craft. The Brand Voice feature helps maintain coherence and identity across writing, ads, emails, and scripts.

• Content Repurposing: The process of transforming existing content into different formats or versions for various platforms. For example, Craft can convert a webinar transcript into a blog post, a Twitter thread, and a promotional video using its multi-tool ecosystem.

• Creative AI: Refers to the use of artificial intelligence in tasks traditionally associated with creativity, writing, designing, composing, and storytelling. Craft is positioned as a leading Creative AI platform due to its breadth and depth of tools.