codebasics

While we deal with uncertainties of life, let’s not forget our true nature.

सर्वस्य चाहं हृदि सन्निविष्टो (Gita ch: 15)

The divine element that powers sun and galaxies resides in our hearts too. Let’s celebrate this Diwali by experiencing that inner light within us.

Wish you all a very shubh dipawali 🪔

1 day ago (edited) | [YT] | 121

codebasics

Hello everybody, Docker is a common skill that AI engineers, FDE and Software engineers need to have. To help everyone learn this, we have published a complete docker course on YouTube. Check it out.

5 days ago | [YT] | 27

codebasics

20 million subscribers in one room.

That was the vibe of a Youtube dev summit in NYC last week. Got an opportunity to meet amazing creators such as,

Corey Schafer
Beau Carnes (Freecodecamp)
Haris Ali Khan (Code with Harry)
Alberta Devor (Alberta Tech)
Thu Hien Vu (Thu Vu)
Lucy Wang (Tech with Lucy)
Sam Witteveen (Sam Witteveen)
Daniel Shiffman (Coding train)
Muhammad (Promp Engineering)
Ania Kubow (Code with Ania Kubow)
Qingyue(Annie) Wang (Google Cloudtech)

Some highlights,

1. Recorded a crazy podcast with Google deepmind engineers Thor Schaeff and Jack Wotherspoon (It will be live soon on our channel)

2. Learnt YT growth lessons from Paddy Galloway a person who has advised big YouTubers, including Mr. beast.

3. The magic show during the night was truly magical 😊 (still trying to figure out how a $ bill got inside a lemon)

4. Networking and insights sharing with other fellow creators

5. Harry's 10 million diamond button

Thanks Google for arranging this. Big kudos to Vivian Jair, Asrar Khan, Greg Baugues, Jen, Isa and everyone else for all the hardwork for making this event a success.

#codebasics #data #ai

Skip the Carousel

1 2 3 4 5 6 7 8 9 10

6 days ago | [YT] | 84

codebasics

Step-by-step study plan for becoming an AI engineer (with a GenAI-first approach)

1) Python:

Learn the fundamentals, working with APIs, handling JSON and writing clean, reusable code. Learn Pydantic for data validation and FastAPI for building APIs.

2) LLM fundamentals:

Understand tokens, context windows, embeddings, prompting, structured outputs and model limitations. Write your first Python program to call an LLM through GroqCloud or the Gemini API. Both offer free tiers for experimentation

3) RAG:

At my company, AtliQ, ~50% of our client projects use RAG. Based on that experience, it’s a skill I recommend prioritising. Learn document processing, chunking, embeddings, vector search and retrieval. Build an application that answers questions using your own data.
Start with Chroma, then explore Qdrant. Learn metadata filtering, hybrid search and reranking as you progress.

4) Tool calling and agents:

Connect models to tools and external APIs. Learn workflows, state management and when a simple workflow is enough. Start with direct API calls, then learn LangChain. Explore LangGraph when you need more control over stateful agent workflows.

5) Evaluation and reliability:

Create test datasets. Evaluate retrieval quality and answer faithfulness, test for hallucinations and prompt injection, and measure cost and latency.
Tools: Ragas for evaluation; LangSmith for tracing, debugging and evaluation. Combine automated checks with human review.

6) Deployment and monitoring:

Learn Git, SQL, Docker and cloud basics. Serve your application using FastAPI and deploy it with logging, monitoring and secure API key management.
Use LangSmith for tracing and production monitoring. Explore LangSmith Deployment if you want managed hosting for your agents.

7) Portfolio:

Build two or three complete applications that solve real problems. Show how they work, how you evaluated them and what you would improve. Record a video with a presentation as if you are presenting to business stakeholders. Post it publically on LinkedIn and other platforms to improve your online credibility. I know people who have got full time job offers by following this approach.

8) Optional deeper track — ML and deep learning:

Study statistics, classical ML, neural networks, PyTorch and transformer architecture when your interests or target roles require deeper model work.

Start building early. You don’t need to finish ML and deep learning before shipping your first useful GenAI application. Your project portfolio and building in public matters much more than your fancy collection of certificates.

1 week ago | [YT] | 74

codebasics

As part of our 75 Free Live Webinars happening from October to December, we're bringing industry experts together to share insights on how AI is transforming different fields.

Whether you want to understand AI's impact on your role, learn something new, or explore a topic you're curious about, there's a session for you.

Here are the webinars for the coming week.

Registration link in the comments.

#codebasics #data #ai

1 week ago (edited) | [YT] | 88

codebasics

My first salary in 2012 was 22K INR

In 2026, the fresher salary is almost the same

Fresher salary is stagnant!

What could be the issue?

[1]Education system failing to make people skilled?
[2]IT companies exploiting freshers?
[3]or it's a crazy multivariate equation?

We don't know.

So you can't fix it this Wednesday.

You have two options, CHOOSE

[1] Complain about things you can't control
[2] Focus on what's in your hand

For those who chose option 2

Do this

Treat yourself as a product and train
until you become a market fit.

When you become Market fit,
You are relevant.

When you are relevant,
You are rewarded.

So

Research
Train
Show
Identify gaps
Retrain

until you become Market Fit.

Who is ready to do this?

#codebasics

1 week ago | [YT] | 157

codebasics

I have worked with people over 20+ countries. People are so different but there are some common traits among people who are high achievers and also great to work with.

Listing them here

1. They are not people pleasures. Nice but not pleasing.

2. They think before they commit, no commitments are made without thinking through.

3. Proactive - None of them will make you follow up for something they commited.

4. Quick learners - they know that they don't know everything and are open to learn on the fly.

5. Clear. They are very clear about what their goals, boundaries and what they need from others.

6. High EQ - they have high emotional quotient. They easily understand what matters to others and it makes it easy for others to work with them.

7. Time - they respect their time the most and do the same with others.

8. Candid - they do not hesitate to call a spade a spade. They know how to give and take constructive feedback.

9. High Integrity - they do the same thing with or without somebody watching them.

10. Fitness - they focus a lot on their fitness. It includes their eating, fitness and wellness routines.

I believe there are more, tell me if you know it in the comments.

Pic: A good memory from the Demand Planning Forum - the annual meet where our wider team working across the world meets together in person.

#codebasics

1 week ago | [YT] | 80

codebasics

Quick question for everyone here 👇
Have you ever thought of building your own tech or AI product?

2 weeks ago | [YT] | 34

codebasics

Here’s what’s coming up next week at Codebasics!

As part of our 75 Free Live Webinars happening from October to December, we’re bringing you sessions led by industry experts across different fields.

From Monday to Saturday, explore topics that match your interests, career goals, or curiosity.

Whether you want to explore a new field, strengthen your skills, or understand where AI is heading, find a session that interests you and join us live!

You can also explore recordings of previous webinars and catch up on sessions you missed.

Registration link in the comments.

#codebasics #data #ai

2 weeks ago | [YT] | 40

codebasics

We completed our teaching cohort last month.

The idea behind this cohort is simple: create better teachers for the world.

We explored the fundamentals of teaching, pedagogy, presentation formats, learner engagement, and how to explain complex ideas in a simple and effective way.

The learners stepped out of their comfort zones, recorded teaching videos, practiced what they learned, received feedback, and built meaningful connections along the way.

Fill out this form in the comments if you are interested in joining the next one starting on 4th November 2026 (there are no fees).

#codebasics

2 weeks ago | [YT] | 94