Project terms,
explained in plain language.

We explain, in a sentence or two, the terms that come up often in public-sector and enterprise IT projects, from AI adoption to system implementation, operations, and contracting. Each term links to a related page.

AI & data

Terms you often hear when considering AI.

  • Enterprise knowledge search (RAG)

    A method that first finds internal documents related to a question, then has AI write an answer based on what it found. It can be set up to show the source documents alongside the answer.

    Related page Enterprise knowledge search (RAG)⁠
  • Large language model (LLM)

    An AI model trained on large amounts of text to understand and generate language. It is used for text-based work such as summarizing, drafting, and answering questions.

    Related page AI solutions⁠
  • Prompt

    The instructions and reference material given to an AI model along with a task. Even with the same model, the results vary depending on how the prompt is written.

    Related page What happens if the AI produces a wrong result?⁠
  • AI agent

    An AI setup that, given a goal, breaks the work into steps and calls permitted business tools to carry it through. The functions it can use and the points where a person approves are set in advance.

    Related page Document & workflow automation⁠
  • Pilot (PoC)

    A validation stage, also called a proof of concept (PoC), that tests what an approach can and can’t do on a small task with limited data before a full build. Based on the results, you decide whether to scale up, refine, or stop.

    Related page First AI project: start with a small pilot⁠
  • Video analytics

    Technology that picks out information such as people, objects, and movement in camera video and turns it into data you can use in your work.

    Related page Video AI & data analytics⁠
  • De-identification

    Processing that removes or replaces the parts of personal information that could identify a specific person, so no one can tell whose information it is. It is usually considered before data is used for analysis or AI validation.

    Related page Do we need to use real business data to validate AI?⁠
  • Key performance indicator (KPI)

    A key metric chosen to judge how close you are to a goal. Even for metrics with the same name, the calculation method and reference date must be defined before the numbers can be trusted and compared.

    Related page Dashboards: define the metrics before the screens⁠
  • Dashboard

    A screen that brings several metrics and status information together in one place. It is organized around the decisions its viewers make, so they can move from a summary down to the details.

    Related page Business solutions⁠

System implementation

Terms that define scope when building or changing a system.

Operations & quality

Terms for discussing quality and operations around launch.

Projects & contracting

Terms used in procurement, contracts, and project stages.

Feel free to ask about any term that needs more explanation.

Tell us about the project you’re considering, and we’ll work with you to outline the next steps for a consultation.