Make.com is a website where you connect apps so a task runs by itself, with no code. You can open a free account tonight and build a small working automation in about 30 minutes, before you pay anyone for an AI automation course.
That first build tells you more than any course page can. It shows you whether you enjoy this kind of work. It also shows you where beginners really get stuck, which is rarely the AI part.
What is Make.com, in plain words?
In Make, one automation is called a scenario. A scenario is a chain of small boxes called modules. Each module does one job in one app.
The first module is the trigger. Make's help pages say a trigger "tracks changes in a service and brings them to Make". A new email, a new form answer or a new row in a sheet can all be triggers. A scenario has only one trigger, and it always sits first.
After the trigger come actions. An action does something with the data, like writing a row or drafting a reply. Make also has search modules, which look up records you already have.
That is the whole idea. Something happens, then a few steps follow. An "AI automation" only means that one of those steps asks an AI model to read or write something.
Make, Zapier or n8n: which should a beginner try first?
All three tools do the same basic job. The difference that matters to a beginner in Nepal is the free plan, because paying from Nepal needs a dollar card.
A free Zapier account stops at two steps. n8n has no free cloud plan, and running it free on your own machine means leaving that machine switched on. Make's free plan gives you 1,000 credits a month and 2 active scenarios, and a scenario can branch into different paths.
So our pick is Make. In our own course, students compare the three tools once, then build fully in Make for exactly this reason.
How fast do the free Make credits run out?
This is the part most people learn too late. Make counts work in credits. Its help page says trigger modules use 1 credit per run, "regardless of whether they return data".
Say your scenario checks Gmail every 15 minutes. That is 96 checks a day, and 96 credits a day, even on a day when no email arrives. At that pace, 1,000 free credits last about ten days.
Two more things use credits. An action uses 1 credit for each item it handles, so writing 10 rows costs 10 credits. An AI step on Make's own AI provider costs 1 credit plus more credits for the amount of text it reads and writes.
The fix is simple. On the free plan, run a scenario by hand when you need it, or once a day. That is how we run every class build.
Why connecting Gmail is the hard part
Once they understand the idea, most people want Make to read their Gmail. This is where beginners get stuck, and it has nothing to do with AI.
To connect a personal Gmail account, Make's own guide has you create a sign-in key in Google Cloud, called an OAuth client. The steps are a Google Cloud project, the right APIs, a consent screen, the client keys, and then the connection in Make. The same guide warns that if your project stays in Testing status, you have to sign in again every week.
None of this is difficult once you have seen it done. It is confusing the first time, because Google Cloud uses words a non-technical person has never met. In our course, the mentor spends 30 to 45 minutes walking every student through this one step.
Where the AI step goes, and why a person still checks it
An AI step is just another action in the chain. It reads an email, writes a short summary or a draft reply, and passes it to the next step.
The risk is that the AI writes something wrong and the next step sends it. We would not let a beginner's automation send anything to a customer without a person reading it first. Our rule is short: AI drafts, you approve, then it sends. Every step that sends, posts or pays waits for that human check.
What a course adds after the free build
If the free build below feels easy and you only want small personal flows, Make's free plan and its help pages can take you a long way. A course is worth it when you want the Google connection set up properly, a business flow built end to end, and someone to check your work.
Saarathi Academy in Old Baneshwor, Kathmandu runs the No-Code AI, Agents and Automation course. It lasts 5 weeks, 2 hours a day, and you never write code. Classes run in person and online, with up to 10 students in the classroom. You build one automation fully in Make, plan and price it like a small job for a business, and run one AI agent errand under supervision. The course ends with a certification exam taken online at home.
Try it tonight: your first Make scenario in 30 minutes
This build needs no Gmail and no Google sign-in. It catches a message from a web link and saves it in Make's own small database, called a data store. The free plan includes 1 MB of storage, which is enough for one data store.
- Sign up for a free account at make.com and create a new scenario.
- Add the first module. Search for Webhooks and pick Custom webhook. Click Add, give it a name and save. Make now gives you a web address and starts listening. A webhook works like a doorbell: when that address is opened, the scenario wakes up.
- Copy the address and add
?name=test&message=helloto the end. Paste it into a new browser tab and press Enter. The page should show the wordAccepted, and Make picks up your two fields. - In the left sidebar, open Data stores and add one. Give its data structure two text columns, name and message.
- Add a second module: Data store, then Add/replace a record. Pick your data store, and map name and message from the webhook into it.
- Click Run once. Open your link again with new words in it, then browse your data store and look for the new record.
Each run uses about 2 credits, one for each module. Now picture a Google Form in place of the browser link, and a sheet in place of the data store. That is the start of a real enquiry log for a shop.
The week-by-week syllabus for the No-Code AI, Agents and Automation course is public here, so you can see every week before you decide.
Top comments (0)