Plain definitions of the ideas we work with every day, written to be useful rather than clever.
- AI-native
- AI-native describes a company or product whose workflows, decisions, data and software are designed so AI is part of how work gets done, not a tool used beside it.
- Agentic workflow
- An agentic workflow is a process in which an AI agent plans and carries out several steps, using tools and data, toward a defined goal, with human checkpoints where needed.
- Forward deployed engineer
- A forward deployed engineer works inside the customer's business, discovering the real workflow, prototyping with users and connecting data and tools so the build stays tied to the outcome.
- Retrieval-augmented generation (RAG)
- Retrieval-augmented generation is a pattern where a system retrieves relevant, approved documents and passes them to an AI model as context, so answers are grounded in your own information.
- Human-in-the-loop
- Human-in-the-loop means a person reviews, approves or corrects an AI system's work at defined points, especially where judgment, accountability or high impact is involved.
- Tool calling
- Tool calling is the ability of an AI model to request a defined function, such as searching a database or booking a meeting, which the application then validates and executes.
- AI orchestration
- AI orchestration is the coordination of models, tools, data sources and people so a multi-step process runs reliably, with routing, retries, logging and handoffs.
- Vector search
- Vector search finds information by meaning rather than exact words, by comparing numerical representations (embeddings) of text.
- Workflow mapping
- Workflow mapping is documenting how work actually moves, including actors, inputs, decisions, exceptions and handoffs, before deciding what to change or automate.