Make.com AI agent - Knowledge base questions

Questions
,
1

:bullseye: What is your goal?

I want to upload my database (which is a csv file containing 75933 values).
My goal is to ask AI to find a exact match / similar items within the values.

:thinking: What is the problem & what have you tried?

Is it possible for the AI agent in Make.com to read such big database ?

the database is attached

:camera_with_flash: Screenshots (scenario flow, module settings, errors)

202506-Spec Database.csv (3.02 MB)

3

You can turn that in an array and use the map() function to find matches. Or use the group by option of an aggregator to group items by matching something.

1 Like
5

Thanks.
But, does Make.com support such big database?

6

Good question. I’ve never tried this large of csvs to be honest. Its over 70K rows right? So you might actually hit a timeout limit trying to process all that. And if you are going to process it with an agent, then you will need to upload the file. Cause putting this in the prompt will definitely cause it to timeout or burn a ton of tokens.

But then if you are downloading the file and uploading it to an agent, there shouldn’t be any issues on Make’s side.

1 Like
7

In my experience, the data store of Make is not sufficient for AI processing of large datasets. I prefer using Supabase or Xano with a vectorized database.

Uploading it as a static file is the easiest solution if the dataset does not change.

Vectorization is the way to go for fast and accurate similarity searches at large volumes in a trusted Postgres db. It also offers advanced querying with traditional SQL. The setup is a bit more complex, but the result is incredibly powerful.

Cheers,
Henk

2 Likes
8

Great thanks for your reply!

Do you have any teaching references that I can learn from?

Thanks

9

My mistake, yes:

10

Great thanks! For me as a beginner, hope they can help me to catch it up :slight_smile:

1 Like