Hey devs,
I’m in a weird spot.
Coding fluency: ~0.5/10 (I know print("hello"), loops, and that’s it)
Tech literacy: 9/10 (I get how systems work — auth, file uploads, JWT, rate limiting, etc.)
Execution: 10/10 — I shipped 3 production-grade apps in 3 months
How?
My internship mentor gave me detailed PDF plans (like a 14-day DocuSign clone).
I fed them to ChatGPT/Gemini, let AI generate code, then:
Ran it
Broke it
Fixed it
Tested with Postman
Deployed on Render/Vercel
Used on mobile + PC
Added bcrypt, JWT expiry, 3-try login lock, env vars
I’m not a “vibe coder” — I’m a prompt engineer with high tech IQ and zero syntax muscle.
The 3 Projects (All Live(now free services expired after 2 months) + GitHub)
InDocuSign- https://github.com/AdityaS-EX/DocuSigner
Bug tracker- https://github.com/AdityaS-EX/bug-tracker-app
Fiverr Clone- https://github.com/AdityaS-EX/freelance-marketplace
Google Drive Clone(i didn't upload as i am in deep sht with this mindset)
All have:
JWT + bcrypt
File upload (Multer + Cloudinary/S3)
Postman collections
Rate limiting
Deployed & tested cross-device
Mentor PDF Plans (I Didn’t Make These)
I followed these step-by-step guides from my mentor:
DocuSign Clone Plan (txt)-https://drive.google.com/file/d/1-Cyx-p4M7OP-IulwbuHisAUgBiJ63gde/view?usp=sharing
Bug Tracker(PDF)-https://drive.google.com/file/d/1EWg43p4SzkpR1XkR8icB33At2RKHVOPF/view?usp=drive_link
Fiverr Clone Plan (txt)- https://drive.google.com/file/d/1GP-vsDjchqCG7_CICReE-kzwywED4eum/view?usp=sharing
My Situation (Be Honest With Me)
Truth Reality I can build anything with AII can’t explain a single line without helpI ship fastI’ll fail a live coding interview in 2 mins
I understand system design I don’t know Flexbox vs GridI want ₹4-whatever LPA job so i need direction to polish my way
I need coding fluency
Roast Me. Help Me. Guide Me.
Senior devs, please answer:
Is my approach cheating? Or smart(It is dumb I know it)?
Are these projects hireable as-is(nope)?
What’s production-ready? What’s trash?
How do I go from 0.5 → 7/10 coder in 30 days?
Should I put this on my resume? How?
Be brutal. I can take it.
Tech Stack (All MERN)
React + Tailwind
Node.js + Express
MongoDB / Supabase
JWT, bcrypt, Multer, PDF-Lib, Stripe, Socket.io
Render / Vercel / Netlify
Postman for API testing
TL;DR
I’m a tech-obsessed beginner who used AI + mentor plans to build 3 real apps.
I ship. I secure. I deploy.
But I can’t code without AI.Help me become a real developer.
P.S. I used ChatGPT to write this post — because my thoughts are scattered and I can’t form clean sentences.
But I built the apps. I debugged. I shipped.
Now I need your help to level up.
Top comments (2)
The Gods of coding understand the fundamental concepts that a language is built on. When I became a network engineer, the OSI model was my ticket to understanding what networking is ... the skill came in when I was able to comprehend everything I did in context of the OSI model (which is fundamental to troubleshooting accurately) ... even programming a Cisco router - having hierarchical structure interwoven throughout its core ... protocols package data in layers ... VPN tunnels are built in layers (some layers stack better than others). Networking exists in layers from the copper to the text on your screen.
Comprehending OOP — what is it and why does it exist? What are valid arguments against it? what languages are OOP and which ones aren't? How can you tell the difference? Polymorphism, Inheritance, Scope, Datatypes, understanding the power in something as simple as a Java Interface, understanding the pure gold in breaking code down into simple elements, knowing when a class is getting too big ... something that can't necessarily be taught, but when you understand the value in writing code that can scale and can be read and understood (hell Ive written code that ive looked at years later and have no friggin clue what the hell I was doing - that was bad coding because that should never happen).
If you can't tell me what a verb, noun, adjective or pronoun is, I don't care if you speak English, I'll never hire you to write for me ... if you can't explain interfaces, objects, inheritance, polymorphism, encapsulation and abstraction, and talk about them in relevant context, I'll never hire you to write code for me regardless of what you've done in the past.
As is true with any discipline, the difference between people who can wield it with skill vs the people who just patch shit together and make stuff work is in the ability to comprehend the fundamental concepts of the discipline.
An "engineer" might be able to muck his way through designing a 4 stroke engine, but the engineer who has an almost intuitive understanding of thermodynamics, fluid mechanics, material science, dynamics ... who understands relativity, calculus, and chemistry ... thats the engineer who's going to design breakthroughs for a company and they often carry the rest of the group who can piece things together, but don't truly understand the WHY of things...
Your success output is proportional to your educational input ... don't fool yourself ... it's the uneducated who say nonsensical things like "an engineering degree is a waste of time, I can do what they can do without all that headache" ... those guys don't actually DO anything at that level even though they believe they do ... knowledge is key and comprehension is absolutely mandatory if you want to truly advance yourself to the next level.
Look for Blogs from senior, respected software engineers ... read what they write about and you'll understand what is important to know ... and also understand that the software devs today that are making SEVEN FIGURE incomes in AI ... I guarantee you aren't script kiddies that fuddled their way through the ranks ...
Don't brag about your accomplishments in prompting an AI to write code for you ... its no different than saying that you basically don't know shit so you hired someone who does for free ... I use AI for a reality check from time to time in the code I write ... I often bounce ideas off of it to gain perspective or to look for ways to make my code more efficient ... because they absolutely can have conversations in context of those deeper programming concepts and can help you understand them as well. They are masters of language ... which makes them an excellent tool for learning.
Simple, turn off the LLM and do it yourself.