Online coding and AI classes in the Tacna region

Think of the free class as a fair trial. The mentor gives a small task in simple English and sees how the learner responds. Where English is still too steep a climb, the mentor says so openly and proposes another route, such as individual lessons or waiting a few months. For a Spanish-language explanation for parents of younger children, see clases de programación para niños, en español; working adults may prefer AI and coding for professionals in Peru.

import csv

WORDS = ("INFORM", "SISTEMAS", "COMPUTA", "SOFTWARE")

with open("programas-universidades.csv", encoding="latin-1", newline="") as f:
    tacna = [r for r in csv.DictReader(f, delimiter="|") if r["DEPARTAMENTO_LOCAL"] == "TACNA"]

hits = [r for r in tacna if any(w in r["DENOMINACION_PROGRAMA"] for w in WORDS)]
print("Programme rows in Tacna:", len(tacna))
print("Computing, by keyword:", len(hits))

Nine looks like an answer. But a keyword rule is a classifier, and every classifier makes mistakes; the question is which ones. Version 2 runs both methods on every row, keeps a tally keyed by the pair of answers, and prints each row where they disagree.

import csv
from collections import Counter

WORDS = ("INFORM", "SISTEMAS", "COMPUTA", "SOFTWARE")

with open("programas-universidades.csv", encoding="latin-1", newline="") as f:
    tacna = [r for r in csv.DictReader(f, delimiter="|") if r["DEPARTAMENTO_LOCAL"] == "TACNA"]

table = Counter()
for r in tacna:
    by_word = any(w in r["DENOMINACION_PROGRAMA"] for w in WORDS)
    by_class = r["NOMBRE_CLASE_PROGRAMA_N2"] == "INFORMÁTICA"
    table[by_word, by_class] += 1
    if by_word != by_class:
        print("Disagree:", r["DENOMINACION_PROGRAMA"][:48], "|", r["NOMBRE_CLASE_PROGRAMA_N2"])

print("Both say computing:", table[True, True])
print("Keyword only:", table[True, False])
print("Class only:", table[False, True])
print("Neither:", table[False, False])

A Counter can be keyed by a tuple, so table[by_word, by_class] keeps four counts at once, one for each combination of True and False. Six rows are computing by both methods. Three are caught only by the keywords: two teacher-training degrees whose names include computación alongside mathematics, which SUNEDU files under education, and a second specialisation in territorial planning and geographic information systems, matched because of the word SISTEMAS. One row is computing only by category: a second specialisation in telecommunications engineering, whose name contains none of the keywords. Neither method is wrong; they answer slightly different questions, and only reading the four disagreements tells you which one fits yours.

Data: SUNEDU, Programas Académicos, ODC-By, rows with DEPARTAMENTO_LOCAL equal to TACNA. The keyword list and every count are ours. A row is one programme at one set of premises, and the list includes postgraduate and second-specialisation programmes, so it is not a count of undergraduate degrees.

1. Ages 6 to 8. Sort a pile of toys into "has wheels" and "is a car". Find a toy that has wheels but is not a car, and one that is a car but is drawn without wheels.
2. Ages 8 to 11, in Scratch. Make a list of ten words and a rule that says "yes" if a word contains "ball". Which words fool it?
3. Ages 11 to 13. Add "TELECOM" to the keyword tuple and run Version 2 again. Which disagreement disappears, and does a new one appear?
4. Teenagers. Restrict both methods to undergraduate rows (TIPO_NIVEL_ACADEMICO equal to PREGRADO). How many computing degrees does each method find now?
5. University and adults. Run the comparison for every department in Peru and rank departments by how often the two methods disagree.

The same thinking applies to AI. A model that labels things is a classifier too, and checking it against a trusted label, then reading the mistakes, is how you learn when to rely on it. Teenagers who want to go further can look at our page on coding classes for teens in Peru.

Fábrica de Programadores, the ministry's Python scheme for older secundaria pupils, launched in Lima Norte in July 2025 for a first cohort of 1,000, and nothing we read places it in Tacna so far. Olympiads are a different story: ONEM 2026 opened with a school round on 8 July 2026 involving over 1.7 million pupils across Peru, and UTEC charges nothing for its computing olympiads. We are not connected with the DRE, SUNEDU, MINEDU, UTEC, either university or any contest, and our problems are our own.

Figures from the ENAHO household survey as published by INEI in the 2026 statistical compendium (volume 1, tables 11.5, 11.6, 11.13), decimal commas turned into points.

Our suggestions. Grade placement follows MINEDU's age at 31 March, so classmates can be nearly a year apart in age.

University students can read coding courses for university students in Peru, and every Peru guide, English and Spanish, is listed on coding and AI classes in Peru.

Per the tz database, Peru (zone America/Lima) has stayed five hours behind UTC without summer time since 1994, and India keeps UTC plus five and a half hours. Neither changes its clocks.

Book in your own time zone and the hour stays fixed through the year. Schools went back for 2026 on 16 March; school-age learners often use the summer months of January to March to get going, while professionals start whenever their work allows.