Search for a command to run...
Create content, debug code, generate user personas, get cloud advice & more with ProMind.ai's AI-powered team of copywriter, software engineer...
I seek out simple and scalable solutions to various challenges with tech
No comments yet. Be the first to comment.
Program to interfaces, not implementations.
ChatGPT is great for brainstorming, but professionals need structured drafts, consistent tone, and version history. Learn when to use ChatGPT vs a document-first AI writing assistant like WriterOS
Today I’m launching ProMind Writer, a focused product built around one simple reality: a huge number of professionals do not “write occasionally.” They write constantly. Proposals, SOPs, policies, PRD
How Every Layer You Add Changes Everything
Many SaaS companies face a common challenge: despite feature-rich products and aggressive marketing, they struggle with high churn rates. The root cause? A classic case of "feedback blindness" – building in isolation without truly understanding user ...
On this page
Hey there, I'm excited to share with you my latest project, promind.ai. I've been working on this app that uses the ChatGPT model to power an AI team consisting of a Copywriter, software engineer, product manager, and cloud consultant. With promind.ai, you can accomplish tasks such as generating social media content, blog posts, emails, website landing page content, debugging code, generating code, creating user personas, getting cloud tips and advice on cost-savings and much more (you are likely to see more minds by the time you have read this post).
The inspiration behind promind.ai came from my desire to automate parts of my personal workflow. I wanted to create a platform that was both easy to use and powerful enough to handle complex tasks. My goal was to build an AI companion that could help users of all levels, from beginners to experienced professionals. That's why I segmented promind.ai into what I like to call minds, enabling users to select the tasks they want to focus on.
Using promind.ai is quite a breeze. Let’s take a quick example, you are a solopreneur who wants to share a LinkedIn post on a recent product you’ve built - promind.ai 😜. As seen in the screenshots below, it’s as easy as heading on to promind.ai, selecting the copywriter in the left Nav, briefly talk about what you want in your post, select the LinkedIn Task and hit Generate and watch the magic happen.
To build promind.ai, I used a variety of technologies, including the SolidJS UI framework, the FastAPI backend Python framework, and PostgresDB which drives the SolidJS dynamic UI. I wanted to try out a UI library outside my usual go-to - React. I chose SolidJS because it's lightweight, fast, and easy to learn, and I appreciate its reactive programming model, which is similar to React with a few minor differences.
However, I did encounter some quirks around the concept of accessors and setters for reactivity in SolidJS. Once I got the hang of it, the reactivity feature of SolidJS became a breeze. For example, in the code snippet from the project below, rather than using the createSignal value (equivalent of React useState) as is, you have to call it as a function to get its actual value - due to the concept of accessors.
import { Component, createSignal } from 'solid-js';
const Post: Component<PostProps> = (props) => {
const [showAlert, setShowAlert] = createSignal(false);
const toggleAlert = () => {
setShowAlert(true);
setTimeout(() => {
setShowAlert(false);
}, 3000);
};
return (
<Box marginY={2}>
<Box>{props.value}</Box>
{showAlert() && (
<Alert severity="success" color="info">
Copied!
</Alert>
)}
<Button
variant="contained"
size="small"
startIcon={<AssignmentIcon />}
onClick={() => {
navigator.clipboard.writeText(
String(props.value).replace(/ /g, ' ')
);
toggleAlert();
}}
sx={{ marginLeft: 'auto' }}
>
Copy
</Button>
</Box>
);
}
I also used FastAPI, which made it a breeze to generate Swagger documentation and client libraries for the API. At its core, are engineered and optimised prompts which take the input specified by the user and use this to make calls to OpenAI models as seen below. The result obtained is then processed and returned to the user.
import openai as openai
from config import getenv
openai.api_key = getenv("OPENAI_KEY")
response = openai.Completion().completion_client.create(
engine=model,
prompt=f"{prompt}:\\n\\n{example}\\n\\nAnswer:",
max_tokens=max_tokens,
temperature=temperature,
user=user,
)
result = response["choices"][0]["text"]
Overall, building promind.ai has been a rewarding and exciting experience, and I'm excited to share it with you. I believe that this platform will be a valuable tool for many users, enabling them to save time and effort while accomplishing complex tasks. In the coming weeks, I want to make it possible for users to build their own minds on ProMind.ai. Think Notion for AI assistants! 🙂
If you'd like to give promind.ai a try, head over to the website, and let me know what you think!