Learning Journey for a Full Stack Polyglot

Greetings, everyone!

The other day, I found an interesting message on a Discord channel:

Senior developer with experience in AI/ML, Blockchain, Web, and Mobile development...

• AI/ML → LLMs, RAG, Computer Vision, NLP, Speech AI, TensorFlow, PyTorch.
• Blockchain → Smart contracts, DeFi, DEX, NFTs, Solana, Ethereum, Polygon.
• Web & Backend → React, Next.js, Node.js, NestJS, FastAPI, Django, GraphQL, Python, C#.
• Mobile → React Native, Expo, Flutter.
• Databases & Cloud → MongoDB, PostgreSQL, Firebase, Supabase, AWS, Azure, Docker, Kubernetes.

... and I instantly decided that I want to be this guy, and be able to build my own projects. I don't know if this person will ever read my posts, but his message was a huge trigger for me. Last night I could not sleep, and since then I could not think of anything else...

My background is software engineering, but I overspecialized in low-level firmware for systems with critical security requirements. Therefore, a long path ahead before I can call myself a "full-stack polyglot"...

I decided to allocate 25 hours a week for learning, and I asked Gemini to build a development plan for me, which reads as follows:

Phase 1: Establish Core Skills (Approx. 4-6 Months)

This phase focuses on acquiring the foundational languages and tools required for the most in-demand areas: Web/Backend and AI/ML.

• Month 1: Modern Python and Backend Fundamentals (Deep dive into Python 3, Introduction to REST APIs and Web Frameworks, Git and GitHub mastery). Project: Create a simple FastAPI backend that allows you to add, view, and delete "tasks" (a To-Do list API).
• Month 2-3: Data Science and AI/ML Foundations (NumPy for array manipulation, Pandas for data cleaning and analysis, Introduction to Supervised/Unsupervised Learning, Introduction to Docker). Project: Use Scikit-learn and a Kaggle dataset to build and evaluate a basic classification model. Containerize your Python environment with Docker.
• Month 4-6: Full-Stack Integration and Deployment (HTML, CSS, JavaScript, Introduction to React, PostgreSQL basics, SQL queries, and connecting it to your FastAPI backend). Project: Build a React frontend that consumes the data from your FastAPI backend (which now uses PostgreSQL). Deploy the entire stack using Supabase or a free tier of AWS/Azure to get Cloud experience.

Phase 2: Specialization and Portfolio Building (6+ Months)

Once the core skills are solid, you can choose where to specialize first, AI/ML (RAG/LLMs) or Blockchain, as both have significant security implications.

• Option A: Focus on AI/ML
  -- LLMs & RAG: Learn to use Hugging Face models. Focus on RAG (Retrieval-Augmented Generation) architectures.
  -- Computer Vision/NLP: Pick one. For CV, explore OpenCV and image classification. For NLP, delve deeper into Transformers and sentiment analysis.
  Project: Build a LLM-powered AI Agent that answers questions based on a set of internal technical documents (a custom RAG system).
• Option B: Focus on Blockchain
  -- Smart Contracts: Learn Solidity (Ethereum/Polygon) or Rust (Solana).
  -- dApps and Web3: Understand how to connect a frontend (React) to a smart contract using a library like web3.js or ethers.js.
  Project: Deploy a simple, audited Smart Contract (e.g., a secure escrow or a simple token) and build a React dApp to interact with it.

I realize this is huge and that it will take time. Many of the technologies mentioned, I only know that they exist, but nothing more... Mastering them sounds like a titanic task, and I know that the plan will have to change while I start to get a better understanding of each item in the list.

On the other hand, I'm glad to discover a #learntocode tribe on Hive.

Hopefully, I can get some more advice for the community.

Nevertheless, I intend to keep a journal of my experience. Perhaps others would like to follow a similar path.

Posted Using INLEO