Open wanghaisheng opened 2 years ago
What we want is someone who can go and watch a youtube video and then create some slides with some bullet points and heading about the topics that were discussed in those videos.
You need to know the current trends on youtube and tiktok and you should also know how to create a PPT or Google Slide.
You will do research work by watching videos on the topic or reading articles and then coming up with easy to follow PPT bullet points and headings for that topic so that our CEO can record the videos fast on those topics.
for example this is the video that we want to recreate:
"-Matthew Hussey YouTube: https://youtu.be/DPJqnNmQxFA
-Tony Robbins Article: https://www.tonyrobbins.com/love-relationships/has-your-trust-been-betrayed/
-Health Line Article: https://www.healthline.com/health/how-to-rebuild-trust#defining-trust
You need to be good in creating structured content which are easy to understand.
Share examples of your script research work that you have done for youtube and tiktok.
The video editing is just merging a screenshot with a face cam layer (based on RSS feed trigger and other triggers) and smart enough to detect scenarios. I am looking for someone who can build the following workflow needs.
WORKFLOW
Screenshot of URL based on new RSS feed results. The screenshot will then be put in column K (for example) of a Google Sheet. (Zapier can do this, so no priority)
Your automation/script then uses an online video editor to merge this screenshot with my already rendered face cam video. In total, I have 5 scenarios, which means 5 different target audiences.
Once the video is rendered with the online video editor I would like to add a second (already rendered and done) video based on the scenario. This should also be part of the script/automation.
Google Sheets - Video Editor API - Final Video
The final video will be uploaded to Drive and automatically uploaded to loom.
NEEDS
RESULTS
"Scenarios" means all videos follow a certain pattern:
EXAMPLE SCENARIO BELOW.
In general, the script should be flexible enough (have a GUI) so that amateurs/no-coder can add more scenarios & scenario modules.
Scenario 1 - Indeed Outreach (every module is pre-rendered except for the intro as it requires a unique screenshot as background)
MODULE 1: Indeed Intro for keyword "Ecommerce Manager"
TRANSITION (loom "pause" button click)
MODULE 2: Skill Demonstration based on use case "indeed job posts matching keyword "Ecommerce Manager"
MODULE 3: Social Proof. Again, based on the detected scenario.
Best regards Joel
P.S.: An API is not necessarily needed but a non-tech GUI is preferred. Running it locally will require manpower that I try to cut. It is intended for commercial use and I understand that some tasks (final step to upload it to loom) require manual intervention. However, the script should deliver an elegant solution for this.
It actually doesnt upload in queue 1 at a time. It actually slows down your uploads if you do 2 videos at a time or more because they upload at the same time. I upload big file sizes and both percentages go up at the same time so i will only do 1 at a time for that reason.
Need to upload more than one for my gaming channel playthroughs where i will upload 2 a day i like to set them up the night before when I go to bed ready for the morning not effecting the internet during the day!! Very handy
How do i get them to instantly publish them having to go through all my drafts one by one is beyond annoying. Your response would be much appreciated.
So if I drag and drop a whole series I’m working on I can schedule them in an order so that they will all upload chronologically?
What if, I upload more than 1000 videos at a time? Is YouTube allow me to do that? If yes, Then how I am going to update the title and description for each video?
Can you suggest any tool which I can use to automate this process of uploading bunch of video with there specific title and description.
When i do it all of them are uploaded but not published. Its all labeled as drafts even tho its processed and i set all videos to be publicly visible.
How do you enable YouTube studio so you can upload more than 6 videos a day?
How can you upload multiple videos as so they can be in the same video, I have five videos in my phone and don’t know how to upload them into the same video.
Can u do a tutorial on how to do it on mobile?
Oh my god. You're kidding me? I've been dragging and dropping ONE video at a time. Now that I know this, its going to save me so much time! Thank you so much for making such a easy and short video. I can't tell you how annoying it is to try and find an answer to something on YouTube from a video that's too long for my time. This will save me that much time and hopefully help my channel grow that much more. Thank you and I'm subscribing to help your channel grow :)
[ { "name": "searching for interview", "system": "You are a tool for filtering out paragraphs from the interview dialogues given by user.", "system_keys": [], "user": "Interview Dialogues:\n{input_txt}\n\nPlease select the rounds containing one of following tags: {tags}. Note that you should ONLY outputs a list of the speaker name, speaking time, tag and reason for each selected round. Do NOT output the content. Each output item should be like \"speaker_name speaking_time: tag, reason\".", "user_keys": [ "input_txt", "tags" ], "pre_filter": true, "post_filter": true, "split_length": 4000, "split_round": 4 }, { "name": "proofreading", "system": "You are a proofreading tool used to improve the wording, grammar, and logical issues in a given text. Note that the output should maintain the original meaning.", "system_keys": [], "user": "{input_txt}\n\n------\nPlease proofread the given text and output the improved version. Note that the output should maintain the original meaning.", "user_keys": [ "input_txt" ], "pre_filter": false, "post_filter": false, "split_length": 10000000, "split_round": 10000 }, { "name": "proofreading for interview", "system": "You are a proofreading tool used to improve the wording, grammar, and logical issues in a given interview record. Note that the output should maintain the original meaning, as well as keeping the speaker's name and interview time unchanged.", "system_keys": [], "user": "{input_txt}\n\n------\nPlease proofread the interview record and output the improved version. Note that the output should maintain the original meaning, as well as keeping the speaker's name and interview time unchanged.", "user_keys": [ "input_txt" ], "pre_filter": true, "post_filter": true, "split_length": 4000, "split_round": 4 }, { "name": "summarization for interview", "system": "You are a text summarization tool used to summarize the meaning of each round of conversation in an interview record.", "system_keys": [], "user": "{input_txt}\n\n------\nPlease summarize the meaning of each round of conversation in an interview record. Note that the output should be concise and contains key information. The output should be like \"speaker_name speaking_time: summarization\"", "user_keys": [ "input_txt" ], "pre_filter": true, "post_filter": true, "split_length": 4000, "split_round": 4 }, { "name": "interview to news", "system": "You are a news writer who writes news articles based on the given summary of interview records.", "system_keys": [], "user": "{input_txt}\n\n------\nPlease write a news article based on the given summary of interview records.", "user_keys": [ "input_txt" ], "pre_filter": true, "post_filter": false, "split_length": 10000000, "split_round": 10000 }, { "name": "interview to twitter", "system": "You are a Twitter author who writes eye-catching tweets based on the given interview records. Note that the number of words in the output MUST be less than 140. The output should be like \"[#TwoSessions world view #Scholars praise China's tourism industry for providing strong assistance to global recovery#] Swaran Singh, a visiting professor at the University of British Columbia in Canada and an Indian scholar, told China News Service that China has the world's largest tourism industry and outbound travel market. Chinese outbound tourism consumption will bring billions of dollars in revenue to the global tourism industry. The recovery of the tourism industry will also help promote China's economic development and add vitality to social development.\"", "system_keys": [], "user": "{input_txt}\n\n------\nPlease writes an eye-catching tweet based on the given interview records. Note that the number of words in the output MUST be less than 140. The output should be like \"[#TwoSessions world view #Scholars praise China's tourism industry for providing strong assistance to global recovery#] Swaran Singh, a visiting professor at the University of British Columbia in Canada and an Indian scholar, told China News Service that China has the world's largest tourism industry and outbound travel market. Chinese outbound tourism consumption will bring billions of dollars in revenue to the global tourism industry. The recovery of the tourism industry will also help promote China's economic development and add vitality to social development.\"", "user_keys": [ "input_txt" ], "pre_filter": true, "post_filter": false, "split_length": 10000000, "split_round": 10000 }, { "name": "interview to weibo", "system": "You are a Weibo author who writes eye-catching tweets based on the given interview records. Note that the number of words in the output MUST be less than 140. The output should be like \"[#TwoSessions world view #Indian scholars pay attention to China's strengthened participation in global governance#] Swaran Singh, a visiting professor at the University of British Columbia in Canada and an Indian scholar, said that China's Belt and Road initiative is moving towards a new stage, paying more attention to building segmented markets and high-quality development projects. China will also strengthen its participation in global governance and contribute to the global governance process, architecture, and international narrative.\"", "system_keys": [], "user": "{input_txt}\n\n------\nPlease write an eye-catching tweet for Weibo based on the given interview records. Note that the number of words in the output MUST be less than 140. The output should be like \"[#TwoSessions world view #Indian scholars pay attention to China's strengthened participation in global governance#] Swaran Singh, a visiting professor at the University of British Columbia in Canada and an Indian scholar, said that China's Belt and Road initiative is moving towards a new stage, paying more attention to building segmented markets and high-quality development projects. China will also strengthen its participation in global governance and contribute to the global governance process, architecture, and international narrative.\"", "user_keys": [ "input_txt" ], "pre_filter": true, "post_filter": false, "split_length": 10000000, "split_round": 10000 } ]
Price is negotiable
There are four components to this project.
First: Downloading Videos
The bot needs to crawl on YT utilizing two different methods
Method 1: Search Function
Bot needs to have queue of tags followed by “compilation” that it needs to search, download every new video it encounters for processing
Method 2: Channel Search
Bot needs to have a list of channels it actively checks for new videos from and prioritize those videos for download.
Second: Storage
Cataloging
Videos need to be stored as uncut if they are new downloads and haven’t been processed yet. f they have been decompiled each clip needs to be stored with the title of the video, clip number in series, and channel of origin
Searching
Database must support three simultaneous search filters being, keyword, channel name, and date
Third: Processing
Decompiling
Bot needs to reliably cut up videos into individual clips
Cant be on a timer, needs to separate clips
Fourth: Editing
Recompiling
Videos should have an internal 1/0 ID if they have been used recently, set marker at 3 months. Should also be positive if the clip has been used without all other clips being used market across key words
Show strong preference to clips that have no been used Ability to randomly attach clips together, with regard to up to three search categories modifying results being, key word(s), channel name, and date, or none at all
For a given category attach clips randomly with a date modifier
Example: Search “Fail”, current date to previous date, recompile