This workflow automatically generates YouTube Short content daily with AI (based on your prompt), creates the video through an external video-generation API, waits until the video is finished, converts it into a file, and automatically uploads it to your YouTube account. You could potentially have a system where you wake up and a new Short has already been created and uploaded.
01Schedule Trigger Node
Starts the automation automatically at a specific time.
For example, you could configure it to run every day at 10 AM, every Monday, or several times per week.
So you don't have to manually start the workflow.
02Generate Prompt Node
This is where AI creates the content instructions for the video.
03Google Gemini Chat Model Node
Gemini can generate something like: "Create a 30-second YouTube Short about 5 interesting facts about space."
The exact output depends on the prompt/instructions you've configured inside this node.
04Structured Output Parser Node
It's connected to the Generate Prompt node. Its job is to make sure Gemini's response follows a specific output format.
05Secret Payload Node
It prepares sensitive information needed for the video-generation service.
This could contain things such as API information, authentication data, secret values and request parameters.
The important idea is that this prepares the information required for the next API requests.
06JWT Node (JSON Web Token)
It's used to authenticate API requests, so the workflow generates authentication information before requesting access to the video-generation service.
07Exchange JWT for Access Token Node
The service responds with an access token.
The access token is then used to make authorized requests to the video-generation API.
08Generate Video Node
It sends the AI-generated information to the video-generation service.
The video-generation service then starts creating the Short.
09Wait For Video Node
This pauses the workflow temporarily, as generating a video takes time.
Without this waiting period, the workflow could immediately try to download a video that hasn't finished yet.
10Check Video Status Node
After waiting, the workflow checks service status.
11Is Video Ready Node
TRUE — the video is finished, and the workflow continues: Is Video Ready → Convert Video to File.
FALSE — the video isn't finished yet. Instead of stopping, the workflow loops back to Wait For Video.
Then: Wait → Check Status → Is Ready? It keeps doing this until the video is ready.
This is one of the most important parts of the workflow.
12Convert Video to File Node
Once the video is ready, it goes to Convert Video to File.
Converts the returned video data into a proper video file/binary file that the next step can use: video data → actual file.
This is necessary because YouTube needs the video file to upload it.
13Upload Video Node
The workflow uploads the finished video to your set YouTube account.
14Error Trigger Node
This is for handling problems.
For example, something might fail during AI generation, authentication, video generation, API requests or file conversion.
The error workflow can be used to handle the failure instead of leaving the automation completely unmanaged.
Facebook Page Chatbot
Click to zoom
How the Workflow runs
This is an AI-powered Facebook Messenger auto-reply workflow. Its job is to receive a message from Facebook, check whether the request is valid, filter the message, let an AI generate a response, check knowledge base, remember the conversation, and send the reply back to Messenger. This automatically responds to customers without you having to manually answer every message.
01Webhook — Webhook To Get Data From Facebook Node
The webhook waits for Facebook to send information to it.
02GET — Respond to Webhook Node
When Facebook initially connects to your webhook, it needs to verify that your webhook belongs to you.
If credentials are correct: True → Credentials Accepted, and the workflow confirms to Facebook: "Yes, this webhook is valid."
If credentials are incorrect: False → stops.
03POST — Filter Message Node
This node determines whether the incoming Facebook event is something your AI should respond to.
For example, it can filter out unwanted events, non-message events, invalid data, and messages that shouldn't trigger the AI.
If the message passes the filter, it goes through the Kept output.
04AI Agent — Auto-Reply Agent Node
The AI Agent takes the incoming message and figures out what response should be sent.
05Get A Document Node
After the message passes the filter, the workflow gets and uses the document as knowledge base for answering questions.
06Google Gemini Chat Model Node
The AI Agent sends the customer's message to Gemini, and Gemini generates the response.
The Agent tells Gemini what needs to be answered, and Gemini produces the actual response.
07Simple Memory Node
The Simple Memory allows the AI to remember parts of the conversation.
08HTTP Request — Send Reply To Facebook Node
This node sends the AI-generated answer back to Facebook Messenger.
AI Job Scraper + Resume Optimizer
Click to zoom
How the Workflow runs
It starts from a job-search request in Slack, searches for matching jobs, processes each job individually, uses AI to tailor the resume, creates a customized resume document, and then prepares an email draft with the job details. The workflow essentially turns job hunting from a manual process into an automated pipeline.
01Slack Trigger Node
The workflow starts when you send a message in Slack requesting jobs. For example, you could type something like "Find remote AI automation jobs."
This message becomes the starting information for the entire workflow.
02Check If The Query Is Valid Node
The AI checks whether your job-search request contains enough useful information.
It determines whether the request can actually be used to search for jobs.
If the request is valid, the workflow continues. If the request is invalid or incomplete, it sends the request to the Invalid Query path.
03Send "Searching For Jobs" Node
If the query is valid, Slack automatically sends you a message saying that the workflow is searching for jobs.
This lets you know that the automation has started working.
04Invalid Query Node
If the job-search request isn't valid, the workflow sends a Slack message explaining that the query is invalid.
05Get All Jobs Node
The workflow sends the search request to a job-search API.
The API searches for job listings based on the information you provided.
It can retrieve information such as the job title, company, description, location, and job link.
06Check Results Node
The workflow checks whether the job search actually returned results.
If jobs were found, the workflow continues. If no jobs were found, the workflow sends a Slack message explaining that the query is invalid or no results were found.
07Split Out Node
The job-search API may return many jobs together as one large list. This node separates that list into individual job items.
This makes it possible to process each job separately.
08Loop Over Items Node
The workflow goes through the jobs one at a time.
This is important because each job needs its own customized resume.
09Get Resume Content Node
The workflow retrieves your existing resume from Google Docs.
This gives the AI the information it needs about your work experience, skills, and professional background.
Think of this as giving the AI your original/master resume.
10Create Resume Content Node
The AI analyzes the job description and compares it with your existing resume, then creates customized resume content that better matches the specific job.
For example, if the job emphasizes Zapier, n8n, APIs, and workflow automation, the AI can prioritize those relevant skills from your existing experience.
The workflow uses an OpenRouter (Grok) Chat Model to power this AI step and a Structured Output Parser to make sure the AI returns the information in the expected format.
11Check If Resume Is Already Created Node
The workflow checks Google Drive to see whether a customized resume for that job already exists.
This helps prevent the automation from creating duplicate resumes.
12Resume Exists? Node
This is the decision point based on the previous Google Drive search.
The workflow determines whether a customized resume already exists, and the correct path is then followed based on that result.
13Wait Node
The workflow pauses for a short period before continuing.
This can be useful when another process needs time to finish or when Google Drive/Google Docs needs time to update.
14Copy Resume Template Node
The workflow copies your original resume template in Google Drive.
Instead of modifying your master resume, it creates a new copy for the specific job. This protects your original resume.
15Wait Node (Second)
The workflow waits again to give Google Drive time to finish creating the copied document.
This helps ensure the new file is ready before the next step tries to modify it.
16Update A Document Node
The workflow updates the copied Google Docs resume and inserts the customized content generated by the AI.
The result is a job-specific resume tailored to that particular position.
17If Node
It determines whether the workflow should continue to the email-drafting step if there is a valid email available from the job search results.
18Create An Email Draft Node
Automatically creates an email draft for the job application.
It also allows you to review and make any necessary changes to the email and attachment before sending them.
19Send Details To Jobs Channel Node
Workflow sends the job information to a Slack channel. This gives you a notification that the job has been processed.
You can then review the job, customized resume, and email draft before actually applying.
It is also connected to the Loop Over Items node to process each job one at a time.