I run two production RDS instances by myself. I never thought I really needed something like this. What I usually did was just check the details on the AWS dashboard. Thanks for building this. I love it.
I used a similar tool called AsciiFlow (asciiflow.com). This Monosketch seems easier to use. The best part is that we can add diagrams directly inside the source code.
This is the first time I have seen a post like this on Hacker News.
And I just launched Writtte yesterday, a writing platform for drafting articles with built-in grammar checking and AI-based refactoring with custom styles, and for the first time, with a better copy/paste mechanism for other platforms like Medium, and Substack :D
Hi, I recently launched a product and things are going slowly, just a few user signups per day, high bounce rate on the home page, etc. I don't want to do paid marketing because I want to learn marketing and grow the product organically.
I'm mainly a developer and I have no actual marketing knowledge. The only thing I'm doing is reading, asking questions on Reddit, and watching videos about marketing, then applying what I learn to see if there are any results.
Since this is a repetitive task, I created a simple AI agent that runs locally. It tells me what to do next, what will likely happen when I do this or that, where I should post about my product, why I should or shouldn't post on certain social media platforms, and things like that. Since I'm reading books about marketing every day, I added a mechanism to manually add important insights based on my project's needs. It runs locally and uses AWS Bedrock models, Nova Lite for simple cases and Claude or Mistral for more specific situations. Currently it's just a simple CLI tool. It uses online searching, scraping, and analyzing data with custom pre loaded data that entered by me.
By following what the agent suggests, I got upvotes on Hacker News for the first time, my Medium articles got views and claps for the first time, and the product is now well listed on Google search results within just a few days. It's slowly improving my marketing knowledge, seo knowledge and workflow.
Then I told a friend about this who is also trying to learn marketing. Somehow he discussed it with his friends, and after a few days, one of his friend's friends said he'd like to buy the agent for around $5K. That's when I realized this could be a helpful tool for other founders who lack marketing knowledge. But now I'm thinking about what to do. Should I sell it to him (but keep using it myself), just use it myself, or make it available to everyone by creating a product around it?
For me, I really don't care about the money in this situation, I care more about learning marketing.
I'd really like to know your opinion on this matter.
After thinking about this for a few hours, I decided to create a simple waitlist to see if others need this. If there's enough interest, I may either open source it or build a reusable AI agent for others.
Yes, I agree with your points. When it comes to using compute minutes instead of an I/O credit system, I do not think providers will adopt that model. Without support from them, application developers cannot implement it this way, because it becomes more costly when applications and providers use different approaches.
I wrote about a practical approach to converting AI token usage into credits without floating-point arithmetic. The key insight is using multi-layered integer scaling (price scaling × credit conversion × per-million-token base) to maintain precision across millions of transactions.
I agree with your points. I can clearly see the difference in myself between how I used to program before and how I do now.
At one point, I was using Copilot very heavily. One day I lost my internet connection and tried to write code manually, but I did not even know what to write. All I had been doing was reading the code and pressing the Tab key.
When I tried to write code myself, my fingers automatically moved to the Tab key with perfect millisecond timing. That was the day I disabled AI auto-completion and decided to use only Ask mode in Copilot.
Now I only use code generation when I 100% (atleast 90%) understand what is happening in the code, or when I am working on something that will not help my future growth, like routine code at my job.
reply