.jpg)
Process description
Automation is an intelligent Telegram bot manager for publishing content on social networks. The user sends a photo or video to a chat with the bot, after which the AI assistant clarifies the details (post caption and social media selection) in dialogue mode, and then automatically publishes the content through the integration with workflow Auto Posting. The bot has a dialogue memory, which allows you to have a natural conversation and use the context of previous messages.
API keys and services
Telegram Bot API - receiving messages and sending responses
Google Gemini 2.5 API - AI model for dialogue
ImgBB API - image hosting
Cloudinary API - video hosting
N8n Call Workflow - Auto Posting workflow call
System architecture by blocks
SECTION 1: GETTING MEDIA FROM THE USER
1.1 Telegram Trigger — Entry Point
Purpose: Receiving messages from a user in the Telegram bot
Settings: Updates message, Webhook for the bot
1.2 Switch - Classification of incoming content
Purpose: Determining the type of content sent (photo, video or text)
Rules: Photo (images), Video (video), Text (text messages)
SECTION 2: IMAGE PROCESSING
2.1 Get a file - Get a photo file
Purpose: Downloading an image file from Telegram
File ID: taken from message.photo or message.document.thumbnail.file_id
2.2 Get URL - Upload to ImgBB
Purpose: Getting a permanent link to an image via ImgBB hosting
URL: https://api.imgbb.com/1/upload
Result: public image URL
SECTION 3: VIDEO PROCESSING
3.1 Get a file1 - Get a video file
Purpose: Downloading a video file from Telegram
3.2 HTTP Request - Uploading to Cloudinary
Purpose: Getting a permanent link to videos through Cloudinary hosting
URL: https://api.cloudinary.com/v1_1/<ВАШ КЛАУД АЙДИ>/video/upload
IMPORTANT!
In the Cloudinary service, you will need to go to setting=>upload=>upload presets and create a new preset there and select singing mode - unsigned
SECTION 4: AI USER DIALOGUE
4.1 AI Agent — Social Media Manager
Purpose: Intelligent dialogue agent for communicating with the user
Algorithm: Step 1 - Ask for a signature, Step 2 - Refine social networks, Step 3 - Find a link in memory, Step 4 - Call Auto Posting
Agent prompt: You are a social media manager (Instagram, Tiktok, X, Facebook)
The user will send you a photo or video that they want to upload.
First, ask him “What is the signature going to be?” , Then keep your signature and give it one of the parameters.
If the user says they don't need a signature, you skip it.
Then, after the question about the signature, specify which social networks we are posting, if the answer is “to everything”, then upload it to all. If any specific social networks are listed, then only upload them
After you have clarified everything, you must use the Auto Posting tool to upload a photo or video and be sure to send a link to the photo/video to the tool.
COMMUNICATION STEPS:
1. YOU ASK WHAT THE SIGNATURE WILL BE
2. SPECIFY WHICH SOCIAL NETWORKS
3. You take a link that you previously received from your memory (use the think tool to think and find the link in your memory) and make a request to the Auto Posting tool in Russian.
4.2 Google Gemini Chat Model - AI model
Model: Gemini 2.5
4.3 Simple Memory - Dialogue memory
Type: Buffer Window Memory
Allows AI to remember previous posts and media links
4.4 Think — A tool for thinking
Allows AI to find the necessary information in the dialogue memory
4.5 Call Auto posting - Publishing tool
Call workflow Auto Posting for actual publishing (child workflow)
SECTION 5: RESPONSE TO THE USER
5.1 Send a text message - Send a response
Sending a message to a user on Telegram with the result of the publication
Required services and their configuration
Telegram Bot: Create a bot via @BotFather, get Bot Token
Google Gemini: Get an API key through Google AI Studio
ImgBB: Register on Imgbb.com, get an API key
Cloudinary: Register on Cloudinary.com, create Upload Preset
Use cases
Case 1: Posting a photo with a caption
The user sends a photo, the bot asks for a signature and social networks, then publishes
Case 2: Unsigned publication
The user sends a video, replies that a signature is not needed, the bot publishes it on all social networks
Case 3: Memory usage
The bot uses the Think tool to find the signature from the previous conversation
Practical value
For content creators: Publish on social media directly from the messenger, the history of all posts in one place
For SMM teams: Centralized management of publications via Telegram, quick response to relevant topics
For small businesses: A simple tool without complicated training, publish from your phone in a minute
The result of the system
This automation is a personal AI assistant to an SMM manager on Telegram. The combination of natural dialogue through Gemini, context memory, and integration with the publishing system creates a convenient interface for social media management. The user gains experience with a real manager who understands the context, asks clarifying questions and makes publications automatically.
Child workflow:
Process description
Automation for publishing photo and video content to social networks using AI query processing. The system receives a command from the user, analyzes it through OpenAI, determines which social networks to use, generates a signature for the publication and automatically uploads media via the Blotato API. The whole process from getting a command to publishing takes a few seconds, fully automating the routine work of an SMM manager.
API keys and services
Blotato API - social media publishing (TikTok, Instagram, Twitter/X)
OpenAI API - GPT-4.1-mini for team analysis
N8n Execute Workflow Trigger — Launch Entry Point
System architecture by blocks
SECTION 1: RECEIVING AND PROCESSING A COMMAND
1.1 When Executed by Another Workflow - Entry Point
Purpose: Getting a command from an external workflow for processing
Settings: Input query parameter (any data type)
An example of an incoming command: “Post a photo with the caption 'Mug' on Tiktok and Instagram. Link to the photo: https://i.ibb.co/m59P2Tz9/file-1.jpg”
1.2 AI Agent — Team Analysis
Purpose: OpenAI analyzes the user's team and extracts structured data
Model: gpt-4.1-mini
Prompt: You will receive a command. Send only a variable with a signature and social networks in Russian as a dry answer without further comments. Also, send a link to the photo/video in a separate variable.
1.3 Structured Output Parser - Parsing a response
Purpose: Converting an AI response into structured JSON
Schema: Signature (text), Link (url), Social networks (array of names)
SECTION 2: ACCOUNT SETUP AND PREPARATION
2.1 Set Accounts - Configuration
Purpose: Setting account IDs and API keys for each social network
Settings: blotato_api_key, instagram_id, tiktok_id, facebook_id, twitter_id, and other platforms
video_caption: taken from an AI response (Signature parameter)
2.2 Upload — Uploading media to Blotato
Purpose: Uploading photos/videos to the Blotato system for further publication
URL: https://backend.blotato.com/v2/media
Result: getting an internal URL in the Blotato system
SECTION 3: SOCIAL MEDIA DISTRIBUTION
3.1 Switch - Routing
Purpose: Determining which social networks to publish on the basis of AI analysis
Rule 1: If the social network array contains Tiktok, send it to TIKTOK
Rule 2: If “Instagram” is contained, send it to Instagram
Rule 3: If you have “Twitter”, send it to X
Section 4: POST TO EVERY NETWORK4.1 TIKTOK/Instagram/X — Publish
Purpose: Creating a post through the Blotato API
URL: https://backend.blotato.com/v2/posts
Body JSON: post with target, content (text, platform, mediaURLS) and accountID settings
In Body:
Required services and their configuration
Blotato: Register on Blotato.com, get an API key, connect social networks
OpenAI: API key with access to GPT-4.1-mini
Use cases
Case 1: Publishing to one network
Login: “Post this photo to Instagram with the caption 'New of the Season'. Link: https://example.com/photo.jpg”
Result: Posting only on Instagram with the specified caption
Case 2: Multiplatform publishing
Login: “Upload the video to all social networks with the text 'See what happened'. Link: https://example.com/video.mp4”
Result: Posting to TikTok, Instagram, and Twitter/X at the same time
Practical value
For SMM managers: Saving time on routine publications, ability to manage through simple commands, multi-account management
For small businesses: An easy way to maintain a social media presence, no need to know each platform's API
For agencies: Scaling publications for multiple clients, centralized management
The result of the system
This automation is an intelligent content publisher that understands natural language and automatically distributes posts across social media. Using AI for team analysis makes the system flexible and convenient, and integration with the Blotato API ensures reliable publishing on all major social platforms.