~/pritam.mentor

Start here: from zero to your first tech job

A free roadmap for engineering, MCA and BCA students. What to learn, in what order, the best free resources for each step, and a project to finish before moving on. No course to buy.

How to use this

  • Finish the foundation first. Then pick one path: web development or AI. Add placements alongside it from pre-final year.
  • One resource per step. Switching between five YouTube playlists is the most common way students stay stuck.
  • Build the checkpoint project. Watching isn't learning. A deployed project on GitHub is what recruiters look at.
  • 1 hour a day beats 8 hours on Sunday. Consistency matters more than speed.
// step 0 · everyone

Foundation

Start here if you're new to coding, or if you've watched tutorials but never built anything on your own.

⏱ 4–6 weeks at 1 hour a day

  1. Learn one language properly: Python

    Variables, conditions, loops, functions, lists and dictionaries, reading files. Type every example yourself; don't just watch.

  2. Use the terminal, Git and GitHub

    Moving around folders, running programs, git add / commit / push, and writing a README. Every project from here goes on GitHub.

  3. Solve small problems every day

    20–30 easy problems: reverse a string, count words, find duplicates. This is where logic gets built.

✓ Checkpoint before moving on: A public GitHub repo with 3 small Python programs (for example a quiz, an expense tracker and a file organiser), each with a README explaining how to run it.
// path A

Web development (full stack)

You like building things people can click on: websites, dashboards, apps.

⏱ 4–6 months

  1. HTML, CSS and JavaScript

    Build pages from scratch, make them responsive, then add interactivity with JavaScript and the DOM.

  2. React

    Components, props, state, forms and fetching data from an API.

  3. Node.js, npm and a database

    Build a REST API with Express, store data in PostgreSQL, add login. Understand what npm install actually does.

  4. Deploy it

    Put the frontend and API online with a real URL. A project nobody can open doesn't count on a resume.

✓ Checkpoint before moving on: One deployed full-stack app with login and a database (for example a notes or expense app), a live URL, and a README with screenshots.
// path B

AI / GenAI engineering

You want to build with LLMs: chatbots, AI assistants, apps that use ChatGPT- or Gemini-style models.

⏱ 4–6 months

  1. Python for data: NumPy and pandas

    Load a CSV, clean it, group it, plot it. This is how you'll prepare data for any AI work.

  2. Machine learning basics

    What training, testing, overfitting and evaluation mean. Train one simple model end to end.

  3. Build with LLM APIs

    Call a model from code, write good prompts, get structured output, handle errors and cost.

  4. RAG: let the AI answer from your documents

    Embeddings, vector search and retrieval, so the model answers from your PDFs or notes instead of guessing.

  5. Deploy it

    Wrap it in a small API or UI and put it online so anyone can try it.

✓ Checkpoint before moving on: A deployed "chat with your notes" app: upload a PDF, ask questions, get answers with the source shown. Live link plus a README explaining how it works.
// path C · run alongside A or B

Placements: DSA, resume and finding jobs

Pre-final and final year. Start this in parallel once your foundation is done; don't wait until the drive is announced.

⏱ 3–4 months, ongoing

  1. DSA by patterns, not by random problems

    Arrays and strings, hashing, two pointers, sliding window, stacks, binary search, then trees and graphs. Learn the pattern, then solve 5–10 problems on it.

  2. Core subjects interviewers ask

    SQL queries and joins, DBMS normalisation, OOP concepts, basics of OS and networking. Be able to explain each in your own words.

  3. A one-page resume that gets read

    Projects first, each with a live link and GitHub link, one line on what you built and one on what you used. No objective statement, no long skill lists you can't defend.

  4. Where to actually find jobs

    Company careers pages, LinkedIn Jobs (filter: Entry level, past week), off-campus drives and your placement cell. Apply within 2 days of a posting, and keep a simple sheet of every application.

  5. Practise interviews out loud

    Solve a problem while explaining your thinking to a friend, with a 30-minute timer. Then swap. Do this weekly.

✓ Checkpoint before moving on: 100 problems across the patterns, a resume someone senior has reviewed, and 20 applications tracked in a sheet with dates and status.

Stuck?

Everyone gets stuck. Read the error message slowly, search it, try for 30 minutes, then ask, with the exact error and what you already tried. For weekly notes on what companies are hiring for, read the blog.