AI Powered Software Development: SDLC, Tools & AI Coding

Hellllloooo, Huh,

Now you think, what type of introduction is this? This is 2026, at least have some modern way of greeting, some good behaviour. Come on, we have AI now, you can do better than “Hello.” 😂

But what if I’m in the Stone Age? I still don’t use what nowadays people use, then I will be called Stone Age, right?

So let’s understand — in coding, don’t be stuck in the Stone Age. Let’s be modern.

So whom are you waiting for? Lets start.

I remember 2 years ago, my friend was saying, “Dekhi bhai, mai to code khud se hi likhunga, bhai tum likho.”

But hum to likhwayenge.

You remember the dialogue of Mirzapur, “Bahubali ma__rte nahi hai, ma_r_ne ka order dete hain.”

So let’s be Bahubali of software development, and say,

“Namaste AI wale code likhte nahi hain, likhwane ka order dete hain.”

In English translation:

“Namaste AI developers don’t write code, they give orders to get it written.”

We know you are full-stack, you know how to write HTML, CSS, JS, Node.js, Python, etc., but here it’s time to be a supervisor.

But how? You will be using what, what is good, we will explore the topics, tools, and everything. We will explore how to leverage AI in all fields — development, testing, etc.

We are a commander now. What gun—sorry, sorry—what model to use where? So are you ready? How is the JOSH? AI sir, hahaha.

What did you say? Whatever, whether it is AI sir or HIGH sir, doesn’t matter. Use AI a lot, but wisely.

Now let’s start how to be in 1%. Here we need to learn by doing, how to write the best prompt, and all. When we will write anything below by your own, but understand, don’t go blindly. Explore what does what and why we have done that this way.

So be curious, okay?

We will also see how AI can mess up. It will not behave how it should behave, so how to tackle those situations, so we will see that too and the solution.

But remember, understand the topic. Whatever we will use, instead of going blindly, understand what we are doing.

Let’s explore the tools. What we use today may evolve tomorrow, so that’s why we said, understand and be curious.

So before moving, let’s discuss the SDLC.

Software devlopment Life cycle ( SDLC)

So this is the way how a product comes out, how it comes out, we need to break the steps:

Ideation → Features → PRD → Architecture → DB → APIs → UI Design → Development → Debugging → Code Review → Testing → Deployment

So first, the idea comes from your own or from that requirement team, and you finalize the features, and if MVP (Minimum Viable Product) is possible or not, there is a word we learn in engineering: Feasibility, that also we check.

Then comes PRD (Product Requirement Document). PRD is nothing but the detailed document where everything is mentioned. So, everything for a good product, one priority is to be able to solve your problem.

These are the core steps to turn an idea into a proper product, so keep in mind, and business uses that.

Okay, then after the document comes Architecture. Where the Architecture diagram tells what exactly, how the flow would be, what would be backend and frontend, what tools we will use, all.

This is called ERD also. ERD = Entity Relationship Diagram, which is mainly used to represent the database entities and their relationships.

Then we finalize the DB and we design DB. Designing a database is a core thing. Properly, we need to design that, that’s why it is a base step in the product lifecycle.

And we develop APIs in the language we decided.

Note: If you do not understand how AI does that work, I mean, search that term.

Then UI design, that is also called wireframe in the initial stage. And then the real design. Remember Figma and all, these are used for this.

And then we develop the product using code. We used to write code, now we will take help from AI, but remember, these are the core steps. Either you write it or AI writes it.

Then we debug, means fixing the problem that came after development, and then the code review. AI can also do that.

Testing the code, and then we do deployment and make it live for all.

Now let’s discuss some FAQs.

1. What tools are we going to use?

So there are so many IDEs. IDE is the place where you code. VS Code is called an IDE (Integrated Development Environment).

So different companies have developed AI code editors/IDEs like Cursor, Codex by OpenAI, Antigravity IDE by Google, Lovable, Claude Code by Anthropic, etc.

So you can use any, but we will try to use Codex by OpenAI. And remember, whatever you use, use the best model wisely. Why am I saying always wisely? I will tell you later.

All models are trained on different parameters, benchmarks, we will discuss that too.

2. Should I buy a premium subscription to tools or use any free limits?

So there are AI models with code editors. There are free limits; otherwise, we need to buy a premium subscription.

To learn, remember, you can drop your Netflix subscription next month, but you will never drop the subscription of any of these code editors.

You don’t believe me? Try buying one.

3. Can I use alternative tools / tech stack?

Yes, you can use any of your choice.

About the tech stack, we will use MERN — MongoDB, Express, React, and Node.js. So you can use any. It’s the AI era. You can write code and write Python code.

4. Adopt the Change

So remember, as we said, tools we use today might not be there tomorrow. The new setting, the new code editor, so be curious and adopt the change.

Think beyond that, so that you can replicate the things easily. Go beyond the model and IDEs, and beyond the tech stack we use.

Remember: Basic fundamentals remain the same.

So that’s all for this one.

Next, we will discuss about Step 1, that is Ideation.

Stay tuned, and hell, don’t watch Mirzapur Season 2, AI is here. Start this now, haha.

Jokes apart, keep entertainment and infotainment both in life.

Zindagi Na Milegi Dobara.

Bye-bye, take care, and never trouble anyone special — you ma, papa.