How to allow users to send consecutive messages to an AI Telegram Agent and answer in one reply

Im currently trying to build an AI Assistant with make.com but I came up over an problem:

The assistant cant answer in one reply to consecutive messages for e.g.
Hello!
How is it going?
How is the weather today?
He will answer in three replies to this, but i want in only one reply

Alright, I have figured out the “partial” solution.

Whenever a new messages comes, the Telegram Webhook receives it and it goes inside my Airtable database, it is then sttored there temporarly.

So all messages are combined into one signle cell, but the problem is:

The scenario will execute 3 times which mean again three replies, I am new to make and thats why i May ask an stupid question
Im happy to listen to all variations and solutions

Ive saw a person that realised that but hadnt answered me yet.

Ive tried many things but still nothing works :sob:

Thanks

1 Like

I still have the same problem too

Welcome to the Make community!

Any scenario that begins with a Webhook module, like Telegram’s Watch new messages, has a webhook queue. This means you can set the scenario to run once every once in a while (instead of immediately), then aggregate+group all the new items in the queue by username/userID, and then respond accordingly to each user that has messaged the bot.

Hope this helps! Let me know if there are any further questions or issues.

@samliew
P.S.: investing some effort into the tutorials in the Make Academy will save you lots of time and frustration using Make!

1 Like

Hello Samliew

Any scenario that begins with a Webhook module, like Telegram’s Watch new messages, has a webhook queue. This means you can set the scenario to run once every once in a while (instead of immediately), then aggregate+group all the new items in the queue by username/userID, and then respond accordingly to each user that has messaged the bot.

Hope this helps! Let me know if there are any further questions or issues.

Can you take a dive deep :folded_hands: