Can a Longueuil teen or adult learn vibe coding and AI agents in English?
Yes. Modern Age Coders teaches vibe coding and AI agents in English, live online, to learners in Longueuil, Quebec. Teens from 13 to 17 take Vibe Coding for Teens; adults take a separate agents course built on Codex and Claude Code, which admits over-18s only. Learn in a class of fifteen to twenty or one to one.
Longueuil sits across the St. Lawrence from Montreal, and most of its public data is published in French. That makes it a good place to learn something every vibe coder meets sooner or later: AI-written code that works perfectly on the examples in the prompt and fails on the way people actually type. Vibe coding is asking a model for code in plain words; agents add a loop in which the model picks a tool, calls it and reads the reply before choosing its next move. On this page an AI-style search over the city's list of sports and recreation equipment returns nothing for jeux d'eau and nothing for petanque, although the file lists 19 and 27. The fix takes a few lines. Finding the need for it takes a habit we teach: write the test queries first, the way real users would type them. Classes are in English; the data is in French; the lesson is about text.
Facts last verified 23 October 2026. Teaching is online; no Longueuil branch is claimed.
Age decides which AI tools a learner may use, and so which course they join. The free class decides the level.

Python, web and AI projects with assistants teens are allowed to use, tested against inputs typed the way real people type them, accents included.
See the syllabus
Python by hand: strings, lists, loops and functions, so a teen understands what an assistant's code does to text before trusting it.
See the syllabus
Coding agents on real repositories for working adults. Anthropic limits Claude Code to users aged 18 and over, so teens are not enrolled.
See the syllabus
How large language models, retrieval and agents work from first principles, including why they handle some languages and spellings worse than others.
See the syllabusBrowse the course atlas for more than one hundred options and use the coding roadmap to check prerequisites.
The four we are known for
These run underneath everything above. Every one is live and online, placed by ability rather than by age, and the first class is free.

Python, web and AI projects where the learner still owns the thinking.
See the syllabus
Automate the work you already do, then let AI carry part of it.
See the syllabus
Train a model, read what it learned, and be able to say why it is wrong.
See the syllabus
Run AI coding agents on real work without losing control of the codebase.
See the syllabusVille de Longueuil publishes its sports and recreation equipment as open data on Données Québec: 644 rows, from play modules to arenas. We wrote the search an assistant typically produces, tested it with queries people actually type, and fixed it. All code and counts are ours, from the file downloaded on 23 October 2026.
| Query as a person types it | What the file actually says | Draft finds | Folded search finds |
|---|---|---|---|
| jeux d'eau | Jeux d’eau, with a curly apostrophe | 0 | 19 |
| petanque | Terrain sportif - Pétanque | 0 | 27 |
| patinoires et sentier glace | Patinoires et sentier glacé | 0 | 47 |
| aréna - olympia | Aréna, two spaces, then - Olympia | 0 | 1 |
| pataugeoire | Pataugeoire | 21 | 21 |
The draft lowercases both sides and looks for the query inside the type and place columns. That is what a quick prompt usually returns, and it passes the one test most people try, a plain word with no accent. It fails on the other four for three different reasons: a curly apostrophe (U+2019) where the user typed a straight one, an accent the user did not type, and a doubled space in the city's own data.
The fix folds both the query and the data the same way before comparing: apostrophes made straight, accents stripped, case and spacing ignored. Here is the whole exercise:# Searching Longueuil's equipment list in French: a naive draft, then a folded search (our teaching example)
import unicodedata
from openpyxl import load_workbook # pip install openpyxl
sheet = load_workbook("equipsportsrecreatifs.xlsx", read_only=True).active
rows = list(sheet.iter_rows(values_only=True))
head, data = rows[0], rows[1:]
TYPE = head.index("Type d'équipement")
PLACE = head.index("Localisation de l'équipement")
def fold(text):
"""Curly apostrophe to straight, accents removed, case and spacing ignored."""
text = unicodedata.normalize("NFKD", (text or "").replace("\u2019", "'"))
text = "".join(ch for ch in text if not unicodedata.combining(ch))
return " ".join(text.lower().split())
def draft_search(query):
return [r for r in data if query.lower() in (r[TYPE] or "").lower() + " " + (r[PLACE] or "").lower()]
def folded_search(query):
return [r for r in data if fold(query) in fold(r[TYPE]) + " " + fold(r[PLACE])]
# The test queries: typed the way people really type them, written down before any code.
QUERIES = ["jeux d'eau", "petanque", "patinoires et sentier glace", "aréna - olympia", "pataugeoire"]
for q in QUERIES:
print(repr(q), "draft", len(draft_search(q)), "folded", len(folded_search(q)))
The five queries were listed before any code existed, typed the way a Longueuil parent would type them on a phone: no accents, straight apostrophes, single spaces. If the draft had been tested only with pataugeoire, it would have looked correct.
Our count finds 23 type values with a curly apostrophe and 8 with a straight one, and 27 place names with stray or doubled spaces. Folding is not cosmetic; on this file it decides whether a search works at all.
Folding can merge values. On this file it turns 261 raw place names into 246, and we checked all 14 merged groups: each is one place written two or three ways, such as Complexe Sportif Longueuil and Complexe sportif Longueuil. Elsewhere accents can separate real words, so test before shipping.
The column headed Secteur holds a postal code on 637 of the 644 rows. On 6 rows it holds a district name instead (Saint-Hubert four times, Maricourt once and Longueuil once), and one postal code, J4P3K3 on OBJECTID 69, has lost its space. An agent told to group equipment by sector would group by postal code without saying so. Reading the values before trusting the header is the check.
The dataset is Ville de Longueuil's, published on Données Québec under Creative Commons Attribution 4.0 (CC BY 4.0), which the portal's licence page describes as allowing reuse même à des fins commerciales, that is, even for commercial purposes (dataset, licence). Source: Ville de Longueuil, Équipements sportifs et récréatifs, CC BY 4.0. The city does not endorse this page or our classes.
Text is where AI-written code fails most quietly, because the failure is an empty result rather than an error message. Our routine for any code that handles names, places or search terms has four steps.
| Step | What the learner does | What it caught in Longueuil |
|---|---|---|
| 1. Collect real inputs | Write test strings the way users type them: no accents, mixed case, odd spacing | Four of five queries failed the draft |
| 2. Read the data's spelling | Count which characters actually appear: curly or straight apostrophes, accents, double spaces | U+2019 in 23 type values |
| 3. Normalise both sides | Apply the same folding to query and data, never only one of them | All four failures fixed |
| 4. Explain the trade-off | Say what the folding could merge by mistake, and test for it | 14 merged place groups, all the same place |
For agents the stakes are higher, because an agent that gets an empty search result will often conclude that the thing does not exist, and say so politely. Anthropic's published advice to agent builders is to look for the simplest solution and add complexity only when it is needed; here the simple solution is a better search tool, not a cleverer agent. In our adult course, Longueuil learners fix the tool, then write the test that would have caught the empty answer.
This is the owner's promise on our own terms: we teach learners to think about a problem and solve it, never to copy an answer, AI answers included. A Secondary 4 student and an adult analyst in Longueuil both finish this exercise the same way, by explaining in their own words why the draft returned nothing.
Each AI company publishes its own age floor. Our reading of those pages on 22 October 2026 is summarised below, and it fixes which course a Longueuil learner may enter.
| Tool | Teens, 13 to 17 | Adults, 18 and over |
|---|---|---|
| GitHub Copilot | Yes: GitHub accounts start at 13, and verified students get Copilot free through GitHub Education | Yes |
| ChatGPT and other OpenAI services | Only once a parent agrees: OpenAI asks for that consent from every user under 18 | Yes |
| Claude and Claude Code | No: Anthropic's consumer terms set a minimum age of 18 | Yes, in the adult course |
| Any assistant during the CCC | No: the contest bans generative AI and editor plug-ins | Not relevant |
Teens are never steered toward tools their age excludes. Before any lesson that would use an OpenAI product, we ask for a parent's agreement, as the company requires.
Adults practise with agents that change files and run commands, always inside a Git project so each change can be inspected and reverted.
These rules belong to the companies and can change. We look again before each new course; check the live terms before you open any account.
Quebec's education system differs from the rest of Canada's in ways that matter for planning. We quote the Quebec government's pages and CEMC, read 22 October 2026; summaries are ours.
In Secondary Cycle Two, Quebec students follow one of three mathematics options, abbreviated CST, TS and S in the English program document, and the ministry states that each of the three still opens the door to a pre-university CEGEP program.
After Secondary 5 comes CEGEP: two full-time years for a pre-university program, three for a technical one, according to the Quebec government, which also notes that universities weigh applicants by their R score.
Quebec's Charter of the French language makes French the province's official language. We teach in English, which suits learners from either school system who are comfortable following a lesson in English.
| Contest | Languages | Date |
|---|---|---|
| Beaver Computing Challenge (no coding background needed) | English only | 9 to 20 November 2026 |
| Canadian Computing Competition, Junior or Senior | English and French | Wednesday 17 February 2027 |
Entry to any CEMC contest is arranged by a school; there is no personal registration. The CCC is offered in French as well as English, which suits Longueuil students in French-language schools; the Beaver Computing Challenge is offered in English only. We prepare students with original problems of our own and point them to CEMC's free archive, and we do not grant credits of any kind. Our Secondary 4 math options page and CEGEP computer science page go further.
Longueuil keeps Eastern time, as Montreal does. Clocks go back on 1 November 2026 and forward on 14 March 2027; India, where our teachers are, does not change.
| Period | Longueuil offset | India time for that class |
|---|---|---|
| Until 1 November 2026 | UTC-4 | 7:00 pm |
| 1 November 2026 to 14 March 2027 | UTC-5 | 8:00 pm |
| From 14 March 2027 | UTC-4 | 7:00 pm |
A Longueuil class stays at the same local time across both changes; our teachers adjust on their side, and we re-send the schedule the week before.
Numbers on the 450, 579 and 354 area codes are accepted in the booking form. Type the ten digits without the 1.
A Saturday morning in Longueuil is a Saturday evening in India, which is why weekend morning groups are easy for us to staff.
The age bands are a guide; the free class shows where each Longueuil learner actually starts.
Strings, lists and functions written by hand, so AI-written text code can be read.
Python for TeensGit & GitHub Course for TeensProjects with permitted assistants, checked with real inputs in more than one language.
Vibe Coding for TeensAI and Machine Learning for TeensGenerative AI, retrieval and full stack building for CEGEP and university students and career changers.
Generative AI CourseVibe Coding for CollegeCoding agents on real repositories, with tools that are tested before agents use them. Adults only.
Codex & Claude CodePython Automation CourseYes, because the AI wrote a search that could not find anything a real user typed.
The draft in the Longueuil example was tidy and plausible, and it returned zero for four ordinary questions. No error, no warning. Someone had to know that an apostrophe can be two different characters, that accents matter to a computer even when they do not matter to a reader, and that city data contains stray spaces. That knowledge comes from having handled text yourself.
So we teach code and judgement together. Teens use assistants their age allows and adults use agents, but every learner writes the test inputs, reads the data's real spelling, and explains the final code without help. That is how AI help becomes something you can depend on.
In a bilingual region, text handling is not an edge case; it is most of the job. The longer argument is in is coding still worth learning in 2026.
Learners sit at their own computer in their own home; the teacher, live on video, shares the screen with them and asks for the reasoning behind each line, in English.
Weekdays or weekends, flexible: choose your batch, or for a very personalised plan choose 1-on-1 at your convenience, weekend mornings or weekdays.
Fifteen to twenty learners of the same age band and level, or a single learner with a teacher, which adults often choose for their own projects.
Teens and adults learn in separate courses, because the companies' terms allow them different tools.
Brossard, Boucherville and Montreal learners are placed in these groups too; all of them keep Eastern time like Longueuil.
Any recent computer with a browser, a headset if possible, and internet that holds a video call. Nothing paid is required for the AI tools in the teen course.
Session one is a genuine lesson that costs nothing. At the end the teacher suggests a level and course, and we never ask for a card number at that point.
No Longueuil office
We have no centre in Longueuil and do not claim one. Teaching is live and online, in English, from India. We are not connected with Ville de Longueuil, Données Québec, the Quebec ministry or CEMC.
Three items, each displayed in your own currency by the price anchor.
A full first lesson at no charge, with a placement at the end.
The monthly fee for a class of fifteen to twenty, eight live lessons a month.
The monthly fee for one-to-one lessons at times you choose.
There is no Longueuil surcharge or discount: the fee matches every other Canadian city. Payment starts after the first lesson, when you have chosen a course and a time.
Rated 4.9 across 547 Google reviews. These are real reviews, reproduced as written.
★★★★★
"The one step solution for my son. Modern Age Coders make learning coding so simple that kids love it. The teachers explain complex concepts clearly with practical exercises and interactive content."
Ria Mukherjee
Parent
★★★★★
"Modern Age Coders has been a game-changer for me. I struggled to grasp IT concepts and coding before joining, but their classes transformed everything. I can now confidently write complex programs with ease."
Samriddha Mondal
Student
★★★★★
"One of the most wonderful education centres out there. Education is not limited to school syllabus but focuses on skill development."
Vansh Agarwal
Student
★★★★★
"My child Dhairya is really enjoying the Modern Age Coders classes. This is his first online class and he eagerly looks forward to it. I can already see his improvement, and the teachers are very cooperative."
Sonam Oswal
Parent of Dhairya
★★★★★
"Modern Age Coders have wonderful teachers who teach in a clear, easy and practical way. The teacher boosts students' confidence and inspires them to learn without hesitation."
Sonu Goyal
Parent
★★★★★
"I highly recommend this computer coding class! The teachers are incredibly knowledgeable and passionate about coding."
Ritu Kedia
Parent
Lesson one opens with an AI-written search that misses obvious results. How a learner hunts for the reason, and the test input they write to prove it, tells us which level and which age track to suggest.
Questions Longueuil families and adult learners ask us most.
Yes, live online with a teacher. Teens aged 13 to 17 take the teen course and adults take separate courses. There is no classroom in Longueuil.
English. The data example uses a French-language city file, and learners handle French text in it, but teaching is in English.
Asking an AI model to write code from a plain description. We treat that as the first draft; testing it with real inputs and explaining it are what we teach.
No. Anthropic's consumer terms require users to be at least 18, so Claude and Claude Code are limited to our adult course and never offered to teens.
No. Without a parent's go-ahead, which OpenAI requires below 18, we leave ChatGPT out. GitHub Copilot is the alternative, free for verified students.
Yes, online. Adults learn agentic coding with Codex and Claude Code on real Git repositories, reviewing and reverting each agent change; teenagers cannot enrol.
The city file writes it with a curly apostrophe, and people type a straight one. The draft compared them as different characters. Folding both sides the same way finds all 19.
Yes. The CCC is available in English and French, according to CEMC. The 2027 sitting is on Wednesday 17 February 2027 through participating schools, and AI help of any kind is banned.
No. We teach alongside school and CEGEP and do not grant credits of any kind; those come from the student's own institution.
Weekdays or weekends, flexible: choose your batch, or for a very personalised plan choose 1-on-1 at your convenience, weekend mornings or weekdays. Times are given in Eastern time, Longueuil's own.
Read vibe coding and AI agents classes in Canada for the whole country; every province is listed on coding classes in Canada. Quebec students can read about CEGEP computer science preparation; parents may want the AI tools age guide, and adults AI agents explained.