The little glossary of software development
This software development glossary explains, in simple words, the terms you hear in meetings with a technical team: algorithm, front-end and back-end, framework, API, database and big data. It also covers the words of developing with AI: AI agent, coding assistant, vibe coding and AI-generated prototype.
:quality(80))
To go from an idea to software, you first need to speak the same language. Not always easy: software development has its own vocabulary.
If a word has ever left you puzzled in a meeting with a technical team, this little glossary is for you. Each term is explained in one sentence, then with an everyday image.
And since AI started helping write code, the vocabulary has grown: vibe coding, coding assistant, AI agent. We've added them at the end.
What is an algorithm?
It's a word you hear used for everything. The Larousse dictionary defines it as "a set of operating rules whose application makes it possible to solve a stated problem by means of a finite number of operations".
An algorithm is like a recipe: a series of instructions that follow one another to reach a result. The cook, here, is the computer: it follows the recipe written by the programmer to the letter. And as in cooking, the possibilities are endless.
Front-end and back-end: what's the difference?
The front-end is the software's storefront: what appears on screen and reacts to your clicks. The back-end is everything that happens behind the scenes, mainly managing the information it sends back to the front-end when the front-end needs it.
Think of a play. The front-end is the sets and the actors, in direct contact with you, the audience. The back-end is the director hidden backstage, script in hand, whispering the lines so the play runs without a hitch.
What is a framework?
A framework (cadriciel, in French) is a ready-to-use working base: it sketches out the main lines of an application and provides the foundation it's built on.
As with an architect's plans, you have to choose the one that meets the project's needs, taking into account the preferences of the team carrying it out. A good base makes the project more efficient and easier to complete.
What is an API?
API stands for Application Programming Interface. It's software that offers one or more functions to other software. It provides the "what" without showing the "how".
The most common example: the "Sign in with Google" button. The application uses Google's API to confirm who you are, without ever seeing how Google does it.
APIs are also what allow an AI agent to act: it's through them that it checks your calendar, your CRM or your accounting software.
What is a database?
Every time a form is submitted on a website or a system remembers you, there's a database behind it.
According to Oracle, a database is an "organized collection of structured information, typically stored electronically in a computer system". Picture a big table: each entry (a car, for example) is a row, and each piece of information about that entry (the year, the colour) is a column.
| Model | Year | Colour |
|---|---|---|
| Sedan | 2019 | Red |
| SUV | 2022 | White |
| Coupe | 2021 | Grey |
Data, information and fact: what's the difference?
3 words that often get mixed up. The Larousse defines them like this: information is an "element of knowledge that can be represented using conventions in order to be stored, processed or communicated"; data is the "conventional representation of information for the purpose of computer processing"; a fact is "what is recognized as certain, indisputable".
Put simply, a piece of data is a raw value, without context, waiting to be turned into information. The number 7, on its own, is data. The 7 of hearts, in a deck of cards, is information. Fact applies less to computing, but it adds a notion of certainty.
What is metadata?
Metadata is "data about data": it describes another piece of data and gives it its context.
Take a photo: its metadata shows the date and the place where it was taken. It's often thanks to metadata that the source of an image can be checked. Same thing for an Earl Grey tea: its origin, its harvest date and its weight are metadata that describe it.
What is big data?
Big data is the continuous flow of a very large volume of varied data (numbers, text, sounds, videos), produced at a frantic pace, among other things when you browse the Web or social media.
It's analyzed to make predictions: in finance, to forecast stock prices; in insurance, to better assess risk; on video platforms and social media, to recommend content to you. It's often described by 5 "V"s.
Volume
The amount of data, which keeps growing.
Velocity
Collecting data in real time.
Variety
The many forms data takes.
Veracity
The reliability and accuracy of the data.
Value
What the data is worth to the business.
Language model, AI assistant and AI agent: the words of AI.
With Claude, ChatGPT, Copilot and Gemini, new terms have been added to the software development vocabulary. Here's what they mean, without jargon.
The difference to remember: an assistant answers a question, an agent carries out a task and the human has the final say.
Language model
The engine behind Claude, ChatGPT, Copilot or Gemini. Trained on huge amounts of text, it predicts the most likely continuation of a text to answer, summarize or write.
AI assistant
The application you chat with, built on a language model. You ask a question, it answers.
AI agent
An assistant you hand a complete task to rather than a question. It checks your tools, works through the steps and comes back to you for approval.
Assistant that cites its sources
An assistant connected to your own documents, which shows where each answer comes from. You can check before relying on it.
Prompt
The written request you give an AI assistant. The more context and examples it gives, the more useful the answer.
Hallucination
A made-up answer, but presented with confidence. That's why you check the sources and why the human has the final say.
Vibe coding, coding assistant and prototype: developing with AI.
Today, development teams write code with the help of AI, and even someone who doesn't program can describe an application in words and see a first screen working. You go from idea to prototype in a few days rather than a few months. This new pace comes with its own words.
Remember: an AI-generated prototype shows the idea, production-ready software makes it last. In between, there's engineering: security, real data, an architecture that holds up as the business grows and maintenance.
Coding assistant
An AI assistant that writes, explains or fixes code, like Claude, ChatGPT, Copilot or Gemini. The developer reviews, chooses and remains responsible for the result.
Vibe coding
Describing the application you want in words and letting AI write the code, without reading it line by line. Handy for trying out an idea quickly; the resulting code has to be reviewed before it's used for real.
AI-generated prototype
A first version of an application, produced in a few days with AI, to see the idea on screen and get reactions from the people who will use it. It shows the "what", not yet the robustness.
Code review
A developer reading through the code before it's used. With AI, it matters even more: generated code can look correct without being correct.
Going into production
The moment the software is used for real, with real data and real users. That's when security, performance and maintenance matter.
Technical debt
The shortcuts taken to go fast, which will have to be reworked one day. Like a debt, it costs more the longer you wait. A prototype built at top speed can contain a lot of it.
What is recursion?
We'll finish with our favourite term: recursion.
It's the principle by which a process, a formula or an algorithm can repeat itself, in principle infinitely, each time building on the previous result. It's mostly seen in mathematics, with the famous Fibonacci sequence or the Koch snowflake.
An everyday example: a product's packaging that shows the same product, with the same packaging, on which you can again see the product, and so on.
Frequently asked questions.
What's the difference between the front-end and the back-end?
The front-end is the visible part of software: what appears on screen and reacts to your clicks. The back-end is the hidden part: it manages the information and sends it back to the front-end when the front-end needs it.
What is an API, in simple words?
Software that offers a function to other software without showing how it does it. The "Sign in with Google" button is the best-known example.
What's the difference between an AI assistant and an AI agent?
An assistant answers a question. An agent carries out a complete task: it checks your tools, works through the steps and asks you to approve the result.
What is vibe coding?
It's describing the application you want in words and letting an AI assistant, like Claude, ChatGPT, Copilot or Gemini, write the code. It's handy for trying out an idea quickly, but the resulting code has to be reviewed before it's used for real.
Is an AI-generated prototype ready for production?
No. It shows the idea on screen in a few days and lets you get reactions to it. To be used for real, it still needs engineering: security, real data, a code review, an architecture that holds up as the business grows and maintenance.
That's it for this glossary, which we update as words change. Keep it handy for your next meeting with a technical team, just in case.
And if a word escapes you in the middle of a meeting, ask. A good team should be able to explain everything to you in simple words.
Read next
An idea in mind?
Let’s chat!
:quality(80))
Business and innovationHow do you build an app for your business?
Business and innovationGetting software implementation right in your business
Business and innovation7 good reasons to entrust your digital project to an outside firm