What are the best coding classes in Manchester?
Manchester is where the stored program began: on 21 June 1948 the University of Manchester's Small-Scale Experimental Machine, the Baby, became the first computer to run a program stored in its own electronic memory. The city it belongs to is unusually young. Census 2021 counted 551,938 residents, with 11.2% aged 20 to 24 against 6.0% across England, and 30.2% of residents aged five and over in full-time education or at school, against 20.4% nationally. We teach Manchester learners aged 6 to 67 live and online, in ability groups of five to ten or one to one, with a weekly hour written in UK time and teachers working from India. The first lesson costs nothing; afterwards a group place is USD 100 a month and one-to-one teaching USD 150.
The River Medlock runs through the middle of Manchester, and the Environment Agency watches it at two points inside the city: Gurney Street in Ancoats and London Road further down. Both gauges report a water level every fifteen minutes, and both sets of readings are free to download. Put them side by side and an obvious question appears: how long does a rise at Gurney Street take to reach London Road? The first thing most people try gives a useless answer. The two level series correlate at 0.84 or better at every delay from none to two hours, because both are dominated by the same slow drift, so the number cannot say which delay is right. Correlate the changes instead, quarter hour by quarter hour, and the answer jumps out: a negative correlation at no delay, rising to a clear peak at 45 to 60 minutes. That difference, between the levels and the changes, is the lesson this city's own river teaches our teenage learners here.
Facts last verified 21 September 2026. Teaching is online; no Manchester branch is claimed.
Chosen by what the learner wants to build rather than by age alone. Every one begins with a free lesson, taught live, and no card details are asked for.

Games and animations in Scratch, including a river that rises and falls and a boat that waits for the water.
See the syllabus
Python from Key Stage 3 to GCSE depth, practised on files the city publishes about itself.
See the syllabus
Two gauges, 2,827 readings each, and the question of how long the Medlock takes to travel from Ancoats to London Road.
See the syllabus
Scripts that pull a live feed, align two series and flag when one starts moving ahead of the other.
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 syllabusFrom the 2021 Census for the Manchester local authority, read through Nomis. Every figure below is for the city council area, not Greater Manchester.
| Measure | Manchester | England |
|---|---|---|
| Usual residents | 551,938 | Not compared |
| Living in households | 530,136 | Not compared |
| Living in communal establishments | 21,802 | Not compared |
| Aged 20 to 24 | 61,887, or 11.2% of residents | 6.0% |
| Schoolchildren and full-time students, of residents aged 5 and over | 156,241, or 30.2% | 20.4% |
On Census day, 30.2% of Manchester residents aged five or over were schoolchildren or full-time students. The figure for England was 20.4%.
11.2% of Manchester residents were aged 20 to 24, against 6.0% across England, in the same Census.
The Census counts 21,802 people in communal establishments, the category that covers places such as student halls and care homes.
A note on the arithmetic. The five-year age table totals 551,937, one fewer than the headline count of 551,938, because published census tables are adjusted independently to protect individuals. The two are never meant to agree to the last person, so this page quotes the headline total and never adds up the bands to make its own.
From the University of Manchester's Department of Computer Science and its own account of its history.
The Small-Scale Experimental Machine, known as the Baby, was designed and built by F.C. Williams and Tom Kilburn and made its first successful run of a program. The university calls it the world's first stored-program electronic digital computer.
Alan Turing was Deputy Director of the university's Computing Machines Laboratory from 1948 to 1954, where he proposed the Turing Test and worked on biological morphogenesis.
The Department of Computer Science was founded in 1964, with Tom Kilburn as its first professor and head. Its first undergraduate intake, in October 1965, was 24 men and four women.
The department lists further firsts after the Baby: the first modern computer with a hardware floating point unit, the first transistor computer and the first computer to use virtual memory. The university today describes itself as having a student population of over 40,000 and 1,000 degree programmes across three faculties. We are not connected with the university; we mention it because a child growing up in Manchester lives beside one of the places where computing was invented, and that is worth knowing before choosing GCSE options.
What the Baby's first program did
Stored programs are the reason every lesson we teach is possible: the instructions live in memory beside the data, and a program can be changed by typing rather than rewiring. The Baby was the first machine to do that. A Manchester learner writing their first Python function is using the idea their city demonstrated first.
Two Environment Agency gauges on the same river, both publishing every fifteen minutes through the free flood-monitoring API. We downloaded both and measured the delay.
| Measure | Figure |
|---|---|
| Readings at each gauge on shared timestamps | 2,827 |
| Missing quarter hours | Seven, at the same moments at both gauges |
| Correlation of the two level series, any delay up to two hours | Between 0.84 and 0.96, never decisive |
| Correlation of the changes, no delay | Minus 0.24 |
| Correlation of the changes, 45 and 60 minutes | 0.51 and 0.53, the clear peak |
| Storm peaks arriving 30 or 45 minutes later downstream | 11 of 12 |
Learners join the two feeds on their timestamps and count what is missing before anything else. Seven quarter hours are absent from both, at the same moments.
Shift one level series against the other and correlate. Every delay scores highly, from 0.84 to 0.96, and the method cannot choose between them.
Take the difference from one reading to the next and repeat. The correlation turns negative at no delay and peaks at 45 to 60 minutes, and that is the travel time.
The largest rise in the window came early on 4 September 2026. Gurney Street peaked at 0.875 metres at 05:15 UK time; London Road peaked at 0.522 metres at 05:45. Across all twelve storms that lifted Gurney Street above 0.4 metres, the downstream peak came 30 or 45 minutes later in eleven of them. The two gauges each measure from their own zero, so we compare their timing, never their heights.
Why do the levels fool you? Both gauges share a slow background: a dry week pulls both down together, a wet one lifts both, and that common drift is correlated at any delay you pick. The quarter-hour changes strip the drift away and leave only the moments when water is actually arriving. It is the same reason economists compare month-on-month changes rather than raw prices, and the same reason an AI tool asked to link two series will often report a spectacular correlation that means nothing.
Learned on the Medlock, used afterwards on website traffic and sales, sensor networks, stock levels and anything two systems report at once.
| Habit | What it looks like | What it prevents |
|---|---|---|
| Join on time first | Keep only timestamps present in both series, and count the ones that are missing | Comparing a reading with a gap |
| Remove the drift | Work with the change from one reading to the next, not the level itself | A high correlation at every lag |
| Scan a range of lags | Compute the correlation at each delay and look for the peak, not the first large number | Picking a delay by habit |
| Check with events | Match individual peaks and measure their gaps by hand | Trusting a single summary number |
| Say what is not comparable | State when two sensors measure from different zeros or scales | Comparing heights that do not share a baseline |
The fourth habit matters most. The correlation of changes points to 45 to 60 minutes; the storm peaks say 30 to 45. Those are not in conflict: a peak is a single moment, while the correlation weighs the whole rise and fall. A good analysis reports both and explains why they differ, rather than choosing the more convenient one.
Two cartoon buckets filling from the same tap at different times, and a program that counts how many seconds apart they overflow.
The real Medlock feeds in Python: join, difference, scan the lags, then defend the travel time against the storm peaks.
A reusable script that watches two live feeds and reports when one has started leading the other.
We are not connected to the Environment Agency, the Met Office, the University of Manchester or Manchester City Council, and no link is implied. Census figures come from the Office for National Statistics through Nomis; the river measurements are ours, taken from the Environment Agency's published readings.
Manchester's public data has a long memory, and some of it has already stopped.
The Medlock gauge at London Road opened on 1 March 1975. Its highest reading on record is 1.802 metres, set on New Year's Day 2025.
The Ancoats gauge opened on 15 October 2009 and set its record on 28 July 2019. On the River Irk, the Collyhurst Weir gauge has reported since 27 April 1995.
The Met Office's public historic record for Ringway, at the airport and 69 metres above sea level, runs from 1946 to October 2004 and ends with the words Site Closed.
A field that is not what it says
Each gauge also publishes a value called highest recent. At London Road it dates from October 2012, and it is lower than the record set in 2025. A name is a label someone chose once; the only way to know what a field means is to check its values against its definition, which is exactly what learners practise on this page.
The bands are a guide. The first free lesson decides where a learner actually starts.
Programs that count, wait and compare, and the discovery that two events can be linked by a delay.
Scratch Coding for KidsCoding for KidsOpening real data files, lining up two columns and spotting when one moves first.
Python and AI for KidsMaths Through CodingJoining feeds on time, differencing, scanning lags and checking the answer against events.
Data Science for TeensPython for TeensAutomating the checks so a changing lead between two signals is noticed as it happens.
Python Automation CourseData and AI Analytics for Non-Programmers (Excel, Sheets, Power BI)Because the obvious correlations run from 0.84 to 0.96, and they mean almost nothing.
Ask an AI tool whether the two Medlock gauges are related and it will correlate the levels and report a figure near 0.9, with confidence. Ask it for the delay and it may pick whichever lag scores highest, which on these readings is barely different from its neighbours. Nothing in that output is false, and nothing in it answers the question. The travel time only appears once the drift is removed, and knowing to remove it is a judgement, not a calculation.
That judgement travels. The same trap catches anyone comparing two sales channels that both rise at Christmas, two websites that both slow on Monday mornings, or two classes whose marks both improve through the year. A learner who has seen the Medlock's levels agree at every delay will ask, the next time a number looks too good, what the two series have in common that is not the thing being measured.
So the reason for a Manchester child to learn to code in 2026 is not only that the city invented the stored program, though it did. It is that checking whether a striking number is actually telling you anything is becoming one of the most useful things a person can do. The longer argument is in why a child should still learn to code in 2026.
Manchester runs from Wythenshawe in the south to Blackley in the north, and a live online lesson reaches all of it without a tram ride.
A learner in Levenshulme and one in Moss Side can share a lesson from their own desks, and nobody waits at a stop in the rain.
Reception, the Key Stages, Year 9 options, GCSEs and A levels keep the names Manchester schools use, and lessons are taught in English.
A proper lesson on a real task, then a clear recommendation of level, course and weekly hour. No card details at any stage.
Five to ten learners at the same level, gathered from Manchester, the rest of the UK and other countries, so every stage finds a workable hour.
Roughly eight a month at a fixed time, with breaks for half term, exams and holidays agreed with the teacher.
India does not change its clocks, so our teachers are four and a half hours ahead of Manchester in summer and five and a half in winter, which puts UK afternoons and evenings inside their working day.
A city full of students, a slot for each
With 30.2% of residents aged five and over in education, Manchester asks for after-school slots and late-evening ones for university students in roughly equal measure. Both fit comfortably into our teachers' day.
Three figures, and nothing hidden behind them.
A full lesson at no cost, ending with a level, a recommended course and a suggested weekly hour.
A month of teaching, usually eight lessons, with five to ten learners of the same ability.
The same monthly rhythm of around eight lessons, with one teacher for one learner.
Fees are set in US dollars, the single rate for every family outside India, so Didsbury and Harpurhey pay exactly the same and there is no pound price list. Nothing is charged until the free lesson has settled a course and an hour, and the pricing page explains pauses, missed lessons and changes of format.
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
The first task is chosen to fit: a Scratch timing game for a younger child, a first script that opens a table, or the two Medlock feeds for a teenager who likes a real puzzle.
What Manchester families ask us most.
Census 2021 counted 551,938 usual residents in the Manchester local authority area, of whom 530,136 lived in households and 21,802 in communal establishments. Greater Manchester, a much larger area, is counted separately.
Yes. In 2021, 11.2% of residents were aged 20 to 24, against 6.0% across England, and 30.2% of residents aged five and over were schoolchildren or full-time students, against 20.4% nationally.
On 21 June 1948 the University of Manchester's Small-Scale Experimental Machine, the Baby, built by F.C. Williams and Tom Kilburn, ran its first program. The university describes it as the world's first stored-program electronic digital computer.
Learners download the Environment Agency's 15-minute readings from two Medlock gauges, Gurney Street and London Road, align them, and find the delay between them by correlating the changes rather than the levels. On our September 2026 readings it came out at 45 to 60 minutes.
Because both gauges share the same slow drift, so the levels correlate at 0.84 or better whatever delay you choose. Correlating the quarter-hour changes removes the drift and reveals a clear peak.
After school, in the evening or at weekends, with the hour fixed in UK time during the free lesson. Our teachers are four and a half hours ahead of Manchester in summer and five and a half in winter.
Yes. Pauses for exam weeks, half term and holidays are agreed with the teacher in advance, and the course is planned around the school year.
No, and none is claimed. All teaching is live and online, with no premises anywhere in the UK. A learner needs a laptop or desktop, working sound and a steady connection, and the number shown is an Indian one.
The first lesson is free. After that a group place is USD 100 a month for two live lessons a week, around eight a month, with five to ten learners, or USD 150 a month for the same schedule one to one. Course, format and hour are agreed before anything is paid.
Five to ten learners, matched on level, pace and what they want to build rather than on age or postcode. If no suitable group meets at a workable hour, one-to-one lessons are offered instead.
Other English cities have pages of their own, including London, Birmingham and Leicester. For how schooling in England shapes what a learner needs, read coding and AI classes in England; for the river-and-data side of this page, see the data science course. Everything in the UK hangs off the UK coding page.