Automation · Operations and reporting
Build a department's higher education report without opening SNIES
How an AI agent builds a Colombian department's enrollment, institution and accredited program report from official SNIES open data, without opening the portal.
How do I pull SNIES enrollment and program figures without going into the portal?
You ask the Met for the report in chat and it answers with official figures from Colombia's Ministry of Education: how much enrollment the department holds and what share of the country that is, how it splits by level, sector and delivery format, how it moved since 2015, which institutions concentrate that demand, and which programs are offered there, separating the ones holding high quality accreditation. Three things are worth knowing before you ask. Live enrollment reaches 2021 (2024 only if the consolidated datasets are loaded, and 2022 and 2023 are nowhere). The department you query is the institution's, not the student's. And the report arrives as an answer in the conversation, not as a file.
Where it applies: Colombia. SNIES is the information system of Colombia's Ministry of Education. Outside the country this recipe does not apply, because there is no equivalent source connected.
What starts it
When the customer writes
Nothing is scheduled. The Met assigned to the channel receives the question in plain language ("how much enrollment does Santander have and who holds it") and answers in the same conversation. Every answer carries its source note and its year coverage, which is exactly what you copy under the table when the figure ends up in a deck.
Nothing to set up. The Met assigned to that channel picks up the conversation and runs the procedure inside it.
Who runs it
Met for education market analysis
Not a named, seeded Met: an internal Met for a planning office, an education marketing team or a consultancy. It works for whoever has to back a decision with the official figure, not for the student deciding where to study.
Channels: WhatsApp · Website chat
What you need connected
- SNIES Research and analyze
5 steps · 5 tools
The procedure, step by step
Each step shows the tool that runs. They belong to the real connector: if one of them stopped existing, this page would not build.
- 1
Set the year and the national frame
The picture of a year returns the national total, the public and private split, the male and female split, the undergraduate and graduate split, and the top ten departments and institutions. It comes first because it is the denominator: saying a department has ninety thousand students means nothing until you know how many the country has. The most recent year in the live query is 2021, and the Met starts there unless you ask for another.
- 2
Go down to the department and split it by what matters
The query groups enrollment by whichever dimension helps (level of study, delivery format, core knowledge area, type of institution) filtered by department, and returns every row with its share of the total. Grouping by year gives the 2015 to 2021 series, which is what separates a market that grew from one that shrank. And here comes the warning that decides whether the report is read correctly: the department is the institution's home, not the student's.
- 3
Disambiguate institutions before comparing them
Searching by name returns each institution with its code, sector, department and enrollment. This is not an optional step. Words like "Nacional", "Antioquia" or "Santander" appear in the names of several different institutions, and a comparison built on the loose name mixes two entities without it showing in the result. With the name resolved, the query is filtered again by each one and only then are the figures comparable.
- 4
See what is offered and which of it is accredited
The program catalog answers what is offered, at which institution, in which municipality, at what level and in what format, and it can also count how many programs sit in each dimension instead of listing them one by one. The accreditation filter looks for high quality, which is not the same as the basic qualified registry. One quirk of the source worth knowing: the program name column is corrupted in the public catalog, so the search by name works over the degree title awarded and over the core knowledge area.
- 5
Fill in the years and variables the live query does not carry
Faculty and 2024 enrollment are not in the live API: they come from the portal's consolidated datasets, which are a separate query and aggregate by sector, department, institution, program, gender, level of study or contract type. Their annual reference figure is the second semester, and adding both semesters there counts the same person twice, which is the exact opposite of what happens in the live query.
Before you start
What you need ready
The Met solves none of these for you. If one is missing, the automation stops there.
-
Nothing for enrollment and programs, because this is open data
The query runs against the public source of the Ministry of Education on datos.gov.co. There is no account to create and no credential to configure, which makes it the cheapest connector in the catalog to try out.
-
The SNIES connector enabled on the Met
It is switched on per workspace among the Met's connectors. Without it the Met answers from memory about Colombian higher education, and that is where the figures that sound right and do not exist come from.
-
The consolidated datasets loaded, if you are going to ask for 2024 or faculty
This is not a credential of yours. The consolidated datasets live in a table Meteor fills by ingesting the SNIES portal files, and if that load has not been run that one tool answers with an error while the other four keep working. Worth confirming before you promise a 2024 figure.
The limits
What this automation does NOT solve
Better said here than discovered halfway through the rollout.
-
It does not cover every year across every variable. Live enrollment reaches 2021 and the consolidated datasets currently carry 2024, so 2022 and 2023 come out of none of the five tools.
-
It does not return applicants, admitted students, first year entrants or graduates. Nationally the Ministry only published enrollment as queryable data, and faculty appears only in the consolidated datasets.
-
It does not say where the student comes from. The department and municipality in the query are the institution's home, so the figure measures where people study, not where they live.
-
It does not return individual student records. Everything is aggregate: no lists, no names and no row per person.
-
It does not cover primary, secondary or vocational training. The scope is higher education, and a question about schools gets answered by saying this source does not have it.
-
It does not measure employment outcomes or graduate salaries. That is a different source from the same Ministry and it is not connected to this recipe.
-
It does not build a file or send an email. The report arrives as an answer in the conversation: accumulating it in a sheet or mailing it monthly is another connector's job, not this one's.
FAQ
Questions about this automation
The live enrollment query covers 2015 to 2021. Faculty and 2024 enrollment come from the consolidated datasets, which are a separate tool. 2022 and 2023 are in neither of them, and the Met says so instead of estimating them.
Almost always the semester. The live query adds both semesters of the year, which is what matches the dashboard's annual figure. The consolidated datasets use the second semester as the annual reference, and adding both there counts the same person twice. You can also ask for a single semester, and then the answer's note says so.
No, and confusing them changes the whole report. The filter looks for high quality accreditation. Practically every active program holds the qualified registry, so filtering by it distinguishes nothing, because it would leave almost the entire catalog inside.
Yes. First the names get resolved through institution search, which returns the code and the enrollment of each one, and then the query is filtered by each institution. Skipping that step is exactly how a comparison that mixes two similar entities gets built.
Yes, enrollment can be grouped and filtered by municipality. What you have to remember is which municipality it is, namely the one where the institution sits. An online university based in Bogota adds all its enrollment to Bogota even though its students are spread over half the country.
Yes. This is open data from the Ministry of Education published on datos.gov.co, and every answer carries the note with the source and the year coverage. That note is exactly what goes under the table, along with the year and the semester used.
Keep going
Other automations built on the same tools
Operations and reporting
Create a weekly Google Ads report in Google Sheets
How do I send a weekly Google Ads report to Google Sheets without duplicate rows?
Social
- Starts:
- when someone asks for it
- Steps:
- 5, with 5 real tools
- Needs:
- Google Ads · Google Workspace
Operations and reporting
Save the weekly Mercado Libre order report to Google Sheets
How do I save a verifiable weekly Mercado Libre order report to Google Sheets?
- Starts:
- when someone asks for it
- Steps:
- 4, with 5 real tools
- Needs:
- Mercado Libre · Google Workspace
Operations and reporting
A weekly Search Console report, with the history kept
How do I get the Search Console report every Monday without building it by hand?
- Starts:
- at a set time
- Steps:
- 6, with 12 real tools
- Needs:
- Google Workspace
Operations and reporting
Answer who did not turn in an assignment, without opening Google Classroom
How do I know who did not turn in an assignment without going course by course in Classroom?
- Starts:
- when the customer writes
- Steps:
- 6, with 11 real tools
- Needs:
- Google Workspace · Meteor CRM
Want this automation running on your own data?
We will show it to you with your accounts connected, not with a canned demo.