Building a Content Publishing Engine from Voice Memos to Video

What Happens When Ideas Ship the Same Day?

I have to say this again because it still hasn't fully sunk in. There are things I did yesterday that I've not only documented the idea for but executed on and shipped. And I won't even remember them until later today, when at a certain time a scheduled task is going to run, data is going to get gathered, and the report is going to land in my Telegram. Part of the report will highlight what's new. And I'll be like, "Oh yeah, I just thought of that yesterday. And here's the data on it already working."

That means I'm doing things today that are going to get executed on today. And that's just the new norm. That's just the way things are now. Every time I do anything with all these agents backing me up, it spins off all kinds of really meaningful work. It's a problem, actually, because there's so much happening that a single voice memo can't cover it all without turning into a sprawling overview. I probably need to start doing these two or three times a day just to keep up.

The Voice Memo Pipeline Is Just the Start

I've automated the editing and publishing of my voice memo transcripts into blog articles. That's a great way to put out lots of content that's not AI slop. It's my actual words, cleaned up in a blog-editor kind of way, not in an AI-rewrites-everything kind of way. The stopping-the-slop part has more to do with not stomping all over real quality information that I'm giving in a distinct voice. Basically, don't over-edit it. Do just enough to make it passable as professional quality, but leave the voice, leave the information.

So that pipeline is working. But where I want to take it next is the ability to give extra love to select pieces of content. Maybe there's an article that's particularly high quality, or covers a more important concept, or one I feel like could be improved. I want to select that article, pull in a few related ones, and uplift it. Maybe add images. If it's talking about something graphical we built, or analytics we can show, I could grab a screenshot and drop it in. Simple stuff that makes a big difference.

Another thing that would be great is an analysis bot that looks at my whole blog, factors in trends from Upwork and social platforms, and drafts me a suggestion. Pick an article and say, "Let's walk through this one on video." Because a lot of times when I'm making video content, I go down rabbit holes, and sometimes a one-minute video has more impact than an hour-long deep dive. If you can keep it pithy, get straight to the point, it lands harder. And right now, while I'm trying to grow my audience and I've got so many engineering tasks in flight, I don't have time to be primarily a content publisher. Short and focused just makes sense all around.

The Video Pipeline

The blog publishing pipeline takes a voice memo transcript and processes it into an article. The video pipeline would do something similar but for video files. Take a single screen recording or walkthrough, and have a pipeline that processes and uploads it to all the relevant channels: YouTube, Facebook, LinkedIn, X, and Upwork too. This stuff needs to go on Upwork. That platform supports video content and it's where potential clients are looking.

Right now, at the end of the blog pipeline, it sends me a Telegram message with copy-paste-ready text formatted for each social platform so each post appears optimally for each channel. But there's no reason it can't just go ahead and load up a draft post so all I have to do is approve it. I probably need API keys and authentication tokens with the right scopes for each platform, but that's so easy to do now. And once I'm authenticated with the social platforms for posting, that same authentication opens the door to tracking the analytics of those posts.

What Does Unified Analytics Across Every Channel Look Like?

This connects back to the CyberWorld admin dashboard. I need to add a section, maybe under analytics or maybe its own dashboard, where we're tracking social posts. Every time I post something, I can pull the Google numbers on what that article is doing for site visits, and then compare that to the performance on Facebook, LinkedIn, X, everything. Picture a line graph with different colored lines for each metric so we can correlate them. When I post an article and share it across socials, I can see all the performance data right there in one place.

Higher performing articles bump to the top. Then we can have conversations about what worked. And if there's video content that relates to a particular text article, I could tag those to be part of the same series so the analytics show together. The CyberWorld dashboard is going to be a unified platform for all these different campaigns.

The Blog as North Star

I already implemented a job that crawls my blog so that I can use the blog itself as a way to inform my team of where we're going next. Everyone working on the site can load up a briefing of what the blog covers, what direction we're heading. When I have proactive agents looking for something to do, they can look at that report and find the next thing to work on.

What I need next is a daily stand-up. When I get up every morning, the blog analysis job should prepare me a report so we can have an internal conversation about aspects that, for whatever reason, we might not want made public. Maybe it's information that won't resonate with a public audience, or maybe it's private strategy that's still relevant to what we're saying publicly. We keep the blog concise and consumable for the public, and keep the internal briefing rich with details the team needs.

Tracking Initiatives Across a Hyper-Dimensional Timeline

One thing I want to add to the CyberWorld landing site is a timeline page. Not just a flat chronological list, but something that captures the evolution of all my initiatives. Start with a super high-level summary with the major events, then have a page that drills down into the whole timeline of changes and updates.

But here's where it gets interesting in my head. Every day we're moving forward in time, and that has a shape. Each initiative is moving forward in terms of growth. And each one represents a position in a broader space. It's kind of like a neural net. You've got all these different initiatives, and you could plot them in their own lane, vectorize them, and see how they relate to each other. They have synergy. We could find a way to measure the overlap, the shape of the overlap. So it's not just a timeline. It's a hyper-dimensional timeline. I don't know exactly what that looks like yet. It might need to be animated to represent the extra dimensions. But the data is already there in the blog posts and the daily reports. It just needs to be surfaced.

What My Team Is Really Good At

One thing I've noticed is that my AI team is really good at crystallizing ideas into terminology. I'll have an idea, and I know it's a good idea, and I know it's a significant thing to execute on, but I don't know what to call it. And my team will spit out the actual professional term for it. These are real terms in professional niches that I just didn't know because I don't have that esoteric knowledge. If it's a good idea, it's been done before most of the time, and in the right professional circles it's common knowledge. They've coined terminology for it.

There was a term that came up in a conversation about all this content publishing automation. Something like "content publishing engine" or similar. I wish I could remember the exact term because it would make a great title. I need to start writing these down the moment they come up. But the concept is clear: take all the data we're collecting, all the insight we're gaining, use the blog as a North Star, and build a system that handles the full lifecycle from idea to published content to performance tracking across every platform. That's what I'm building.

I also want to arrange group conversations with different agents responsible for different specialties. Pull in the SEO auditor, the Upwork job scanner, the publishing pipeline. Have them all vibe on how to measure and track progress, factor in blog posts, analytics, and trends. Each agent brings a different perspective, and the conversations that come out of that are where the real breakthroughs happen.