DEV Community: Siddhartha Kunwar

Claude Projects: The AI Workspace That Never Forgets

Siddhartha Kunwar — Fri, 03 Apr 2026 14:30:00 +0000

📁 Stop Re-explaining Yourself

Imagine having an assistant who remembers exactly how you like things done—your writing style, your brand voice, and your specific data—without you having to explain it every single time.

That is the power of Claude Projects.

😤 The Problem: "AI Amnesia"

Standard AI chats have a common flaw: Every time you start a new chat, the AI forgets who you are. You have to re-upload your brand guidelines, re-explain your business goals, and re-paste your formatting requirements. It’s frustrating and kills your productivity.

Projects fix this. They act as a dedicated "brain" for a specific goal, where your instructions and files stay saved permanently.

🛠️ How Projects Work

Think of a Project as a high-IQ digital folder. Inside, you can store:

📋 Custom Instructions: Tell Claude exactly how to behave (e.g., "Always use a friendly tone" or "Never use bullet points").
📎 Knowledge Base: Upload PDFs, transcripts, code, or data that Claude should always reference.
💬 Persistent Context: Every chat inside that project "knows" what's in those files.

🎬 Real Example: The YouTube Workflow

A content creator using Claude Projects for "YouTube Metadata" set up his workspace once:

Instructions: "Format descriptions with a hook, 3 links, and my social handles."
Files: Uploaded a template of his previous best-performing descriptions.

The Result: Now, he just drops in a new video transcript, and Claude instantly generates a perfectly formatted description and timestamps in seconds. He hasn't repeated an instruction in months.

💡 5 Project Ideas to Get You Started

Project Name	What It Does
Email Assistant 📧	Drafts replies in your unique "voice" and tone.
Report Writer 📊	Generates weekly summaries from your raw data files.
Social Media Guru 📱	Creates posts based on your specific brand guidelines.
Meeting Master 🤝	Converts messy notes into clean action items.
Coding Buddy 💻	Remembers your specific tech stack and naming rules.

🚀 How to Create Your First Project

Locate Projects: On the left sidebar in Claude, click "Projects".
New Project: Click the "New Project" button.
Name & Describe: Give it a clear title (e.g., "Marketing Copy").
Add Instructions: In the "Project Instructions" box, explain your preferences.
Upload Content: Drag and drop any "source of truth" files Claude needs.
Start Chatting: Every message in this view will now follow your rules!

✅ Key Takeaway

Claude Projects turn a "general AI" into a specialized tool built just for you. If you find yourself copy-pasting the same instructions more than twice, it belongs in a Project.

This is **Blog 4* of a 9-part Claude AI series for beginners.*

Next up: Artifacts — How Claude builds websites, docs, and apps for you (zero coding required!)

Claude's Extended Thinking: The AI That Thinks Before It Speaks

Siddhartha Kunwar — Tue, 31 Mar 2026 18:30:00 +0000

🧠 The Power of "Slow" AI

Most AI tools give you instant answers. But fast isn't always good. Claude has a special mode called "Extended Thinking" where it slows down and works through complex problems step-by-step—like a thoughtful expert rather than a quick guesser.

💡 What Is Extended Thinking?

Extended Thinking is a setting where instead of answering immediately, Claude uses a "reasoning block" to plan its response.

It’s the difference between:

Standard AI: Blurting out an answer based on patterns.
Extended Thinking: Pausing, reading the nuance, checking for errors, and then responding.

How to activate it: Switch the model setting to "Extended" or toggle the "Thinking" button (available in Claude 3.7 Sonnet and above) before sending your message.

⏰ When Should You Use It?

You don't need this to write a "Thank You" email. Save the extra "brain power" for:

📄 Deep Analysis: Summarizing 100+ page legal or technical docs.
🎯 High Accuracy: Generating timestamps for long-form video (4+ hours).
📊 Logic & Math: Analyzing business data or complex coding bugs.
🤔 Strategic Planning: Creating multi-step project roadmaps.

🎬 Real-World Example: The YouTube Timestamp Challenge

Imagine you've recorded a 5-hour video. Doing manual timestamps would take you all afternoon.

With Extended Thinking, Claude can:

Scan the entire transcript for context.
Identify natural chapter breaks (not just random intervals).
Cross-reference mentions of specific keywords.
Format a perfect list: 0:00 - Intro, 15:30 - The Secret Sauce, etc.

The result? About 90% of the work is done in minutes, leaving you to just do a final review.

🧪 What Does It Look Like?

When this mode is active, you'll see a "Thinking" dropdown. If you click it, you can see Claude's internal monologue in real-time:

"I need to verify if the user wants the timestamps in 10-minute increments..." > "Scanning the section on 'Market Trends'..." > "Ensuring the tone matches the previous chapters..."

This transparency builds trust—you can see exactly why it's giving you that specific answer.

🛠 Quick Toggle Guide

Task	Mode Recommended
Quick Email / Text fix	Standard (Fast)
Brainstorming Ideas	Standard (Creative)
Complex Coding / Debugging	Extended (Precise)
Data Extraction from PDFs	Extended (Thorough)

✅ Key Takeaway

Extended Thinking is Claude's "Reasoning Mode." Use it when accuracy matters more than speed. It’s better to wait 30 seconds for a perfect answer than to spend 10 minutes fixing a fast, wrong one.

This is **Blog 3* of a 9-part Claude AI series for beginners.*

Next up: Claude Projects — Your personal AI workspace that remembers everything.

Why Claude? 3 Reasons it Beats ChatGPT and Gemini for Daily Tasks

Siddhartha Kunwar — Tue, 31 Mar 2026 14:22:51 +0000

Why Choose Claude? 🤖

There are many AI tools out there. So why would you choose Claude over something you already know, like ChatGPT or Google's Gemini? Let's break it down with real examples that make sense for everyday use.

🙋 1. The Big Difference: Claude Asks Questions First

Most AI tools will give you an answer right away—even if they don't have enough information. This often leads to "hallucinations" (made-up responses).

Claude does something different. If it doesn't have enough context, it stops to ask you questions. This leads to more accurate, personalized results.

📝 Example: Writing a Welcome Email

Instead of guessing, Claude might ask:

What's the instructor's name?

What tone do you want—friendly or professional?

Are there specific links or dates to include?

This one small habit leads to dramatically better output.

🎨 2. Better Presentation (Artifacts)

When Claude writes an email or a document, it doesn't just dump text on the screen. Using a feature called Artifacts, it creates a dedicated window on the side of your screen to show you:

Professional formatting
Code snippets you can run
Documents that look like real files

No copying. No messy text blocks. No hassle.

🧠 3. Better at Complex Tasks

In real-world tests, Claude consistently handles large amounts of data better than its peers.

The Timestamp Test:
A creator tried to generate timestamps for a 5.5-hour YouTube video.

Other Tools: Produced inaccurate times or crashed.
Claude: Read the entire transcript, understood the context, and generated perfect timestamps in minutes.

📧 The 'Send via Gmail' Power Move

Here is a small but powerful workflow tip. After Claude drafts an email:

Click the "Send via Gmail" button.
Gmail opens automatically.
The subject and body are already filled in.

Note: This usually requires a browser extension or specific integration, but it's a game-changer for workflow.

🏆 Quick Comparison

Feature	Claude	ChatGPT	Gemini
Asks clarifying questions ✅	Yes	Rarely ❌	Rarely ❌
One-click Gmail sending ✅	Yes	No ❌	Yes
Complex task reasoning ✅	Excellent	Good ⚠️	Good ⚠️
Built-in Artifacts ✅	Yes	Limited ⚠️	No ❌

✅ Key Takeaway

Claude stands out because it asks the right questions, thinks more carefully, and makes it easier to act on its outputs.

This is Blog 2 of a 9-part Claude AI series for beginners.

Next up: Extended Thinking — AI that actually "thinks" before it answers.

The 7 GitHub Repos That Will Actually Get You Hired

Siddhartha Kunwar — Sat, 21 Mar 2026 12:55:29 +0000

Let's be honest: the developer job hunt is brutal.

You're told to "learn to code," "build projects," and "grind LeetCode." But where do you even start? It's easy to waste months on tutorials that don't matter, projects that don't impress, and interview questions that feel designed to trick you.

What if you had a clear, step-by-step toolkit for every single stage of the process?

I found an incredible video from SWErikCodes that lays out exactly that: 7 GitHub repositories that form a complete roadmap from learner to hired engineer.

Here is the complete breakdown of this "get hired" toolkit.

Part 1: Learning the Skills & Building Projects

You can't get a job without the skills. These repos will take you from fundamentals to a portfolio-ready developer.

1. Developer Roadmaps

This is your north star. Before you waste time on a random tutorial, this repo gives you visual, step-by-step roadmaps for Frontend, Backend, AI, and more. It stops you from learning the wrong things.

Link: github.com/kamranahmedse/developer-roadmap

2. Project-Based Learning

The single best way to learn is by building. This repository is a massive, curated list of tutorials that guide you through building real-world projects—not just simple to-do apps.

Link: github.com/practical-tutorials/project-based-learning

3. Build Your Own X

Want to really impress an interviewer? This repo has tutorials for building your own database, operating system, or even a blockchain from scratch. It proves you understand what's "under the hood."

Link: github.com/codecrafters-io/build-your-own-x

Part 2: Acing the Technical Interview

You've built the projects, and you've landed the interview. These repos will get you ready.

4. Coding Interview University

Stop randomly solving LeetCode problems. This is a complete, structured computer science study plan. It gives you a university-level curriculum to fill in your knowledge gaps and study smart.

Link: github.com/jwasham/coding-interview-university

5. System Design Primer

System design interviews can be the most intimidating part of the process. This primer breaks down complex topics (like designing scalable systems) into understandable concepts with clear diagrams.

Link: github.com/donnemartin/system-design-primer

Part 3: Finding the Right Job for You

You have the skills and you're prepped. Now, where do you apply?

6. Internship Lists (e.g., Summer 2025)

For students or recent grads, internships are the best stepping stone. This repo is a massive, actively-updated list of internship opportunities with direct links to apply. It saves you countless hours of searching.

Link: github.com/SimplifyJobs/Summer2025-Internships

7. Hiring Without Whiteboards

This is a life-saver. If you're a great developer but not a "whiteboard performer," this list is for you. It's a huge list of companies that don't use generic, high-pressure whiteboard problems in their interviews.

Link: github.com/poteto/hiring-without-whiteboards

🔥 Bonus Round: 10 More Secret Repos

As I was putting this together, I found another video that's too good not to share. Once you've bookmarked the 7 above, this video on "10 secret Github repositories every programmer must know" is your next level. It's packed with amazing finds for public APIs, free programming books, and more.

Check it out here: https://www.youtube.com/watch?v=a1ehrFHAX7g

Now, I Want to Hear From You

Stop trying to figure this all out on your own. Bookmark this post, pick one repo, and get started today.

To get the conversation going, drop a comment below:

Which of these repositories are you going to use first?
What's your #1 favorite GitHub repo for job hunting that's not on this list?

Let's build a master list in the comments!

Why AI "Solving Coding" Might Actually Be the Industry's Next Nightmare

Siddhartha Kunwar — Sat, 21 Mar 2026 12:51:58 +0000

The buzz in 2026 is everywhere: "Coding is solved."

We’re hearing it from AI CEOs and VCs, even while their own projects are sitting on thousands of open bugs. The dream sounds amazing: autonomous agents doing the heavy lifting, a 200% jump in productivity, and a world where everyone is a "software creator" without needing to know a single line of syntax.

But let’s take a second for the reality check the industry actually needs. The real bottleneck in software has never been about how many lines of code you can churn out in 30 seconds.

The Hard Truth: Writing Syntax is the Easy Part

If you’re just starting your career, please don’t be fooled by the hype. Despite the mythology, writing code is actually the easy bit — syntax is just mechanical.

The real "value-add" of a Software Engineer in the age of AI is much deeper:

Modeling the problem accurately to ensure the solution actually works.
Defining the boundaries between different systems for better stability.
Understanding the data flow and how everything connects under the hood.
Anticipating failure modes before they crash your production environment.
Designing systems that can actually survive change and scale.

When Technical Discipline Collapses

There’s a fascinating point about the hidden value of constraints. Historically, building software was slow and expensive. That "hostile environment" actually helped us; it forced teams to be incredibly disciplined. We debated every trade-off, prioritized polish, and killed mediocre ideas early because we couldn't afford to waste time on them.

When AI makes implementation feel "free," that discipline starts to crumble. If a feature only takes 10 minutes to generate, the temptation is to just ship it. But the result isn't better products; it’s a pile-up of "half-baked ideas" that become an absolute nightmare to maintain. The app stores are already full of dead apps — AI is just helping us build them faster.

The Risk of Outsourcing Your "Mental Simulation"

This is the most important part for new devs: When you write code yourself, you’re forced to mentally simulate the system. You have to understand why a function exists, what it assumes, and what happens when things go wrong.

That mental simulation is your training. It’s how you build the "mental muscles" to handle complex problems. If you outsource all that thinking to an AI, you might be closing tickets, but you aren't growing as an engineer. You’re just babysitting "slop code" generated by a machine.

The Big Picture: More Software = More Complexity

A lot of people think AI means we'll need fewer developers, but history suggests the opposite. This is known as the Jevons Paradox.

As it becomes more efficient to create software, we don’t use less of it—we find infinite new ways to use more of it. Industries that used to live in spreadsheets are moving to custom dashboards and automated tools. This "expanding surface area" of software doesn't replace the need for humans; it actually multiplies the complexity that only skilled engineers can manage.

The takeaway? Don't just use AI to save energy. Use it to be 10x more effective—but only after you’ve put in the work to build the mental model yourself.

What's your take?

Are you seeing AI tools help your team stay disciplined, or are they just leading to faster, messier builds? I'd love to hear your experience in the comments!

🚀 Starting My Tech Journey – Why I’m Sharing It Publicly

Siddhartha Kunwar — Wed, 25 Feb 2026 13:19:07 +0000

Hey Dev Community 👋,

My name is Siddhartha, and I'm a Computer Science student currently navigating the vast and exciting world of technology. For a while now, I've been diving deep into various areas: building robust applications with Full Stack Development, exploring the cutting edge of AI/ML, and fortifying digital defenses with Cybersecurity.

Today marks a significant milestone: I'm officially taking my learning journey public! You might be thinking, "Why bother sharing every step of the way?" And that's a fair question. My decision to share my progress, challenges, and insights openly on platforms like Dev.to stems from a few core beliefs and motivations:

Why I'm Choosing to Learn in Public 📝

Accountability & Consistency: Let's be honest, staying disciplined can be tough! By committing to share my progress, I'm setting a public standard for myself. This external accountability is a powerful motivator to keep building, learning, and pushing forward, even on days when motivation wanes.

Connecting & Growing with the Community: The tech community is incredibly vibrant and supportive. I believe that true growth happens through collaboration and shared knowledge. I'm eager to connect with seasoned developers, fellow students, and anyone passionate about technology. Your feedback, insights, and shared experiences will be invaluable.

Inspiring & Helping Others: I remember how overwhelming it felt when I first started, sifting through endless tutorials and feeling like everyone else had it all figured out. My hope is that by openly documenting my journey – the successes, the bugs, and the "aha!" moments – I can offer a realistic and relatable perspective for others who might be just starting out. If my struggles or solutions can help even one person, it's a win!

Reinforcing My Own Learning (The Feynman Technique in Action): Explaining a concept clearly to others is one of the best ways to truly understand it yourself. By structuring my thoughts into blog posts, I'll be forced to articulate my understanding, identify gaps in my knowledge, and solidify what I've learned.

What You Can Expect From My Journey Here 🚀

My goal is to provide regular, insightful content that covers the breadth of my learning. Here's a sneak peek at what you can expect in the coming weeks and months:

Weekly Learning Logs: I'll share what I've been focusing on, be it mastering a new React hook, implementing a Django REST API, experimenting with Next.js for server-side rendering, or diving into new AI models and libraries. I'll highlight key concepts, resources I found useful, and any stumbling blocks.

Project Breakdowns – From Idea to Execution: I'll be working on various projects, and you'll get a front-row seat. I'll cover:

The Idea & Planning: How I brainstormed, planned the architecture, and chose the tech stack.

Implementation & Challenges: The actual coding process, the frustrating bugs, the problem-solving approaches, and the iterative improvements.

Lessons Learned: What went well, what could have been better, and the key takeaways that will inform future projects.

Hackathon Experiences: I love the energy of hackathons! I'll share my experiences, the projects I build under pressure, and the collaborative learning that happens.

Open-Source Contributions: As I gain confidence, I aim to contribute to open-source projects. I'll document my attempts, from finding suitable issues to submitting my first pull requests.

DSA & LeetCode Insights: While primarily focused on practical application, I'll occasionally share thoughts on particularly interesting Data Structures & Algorithms problems I've tackled, especially how they might relate to real-world development scenarios.

Cybersecurity Deep Dives: Exploring topics like web vulnerabilities, secure coding practices, and basic ethical hacking exercises.

My Philosophy: Show the Journey, Not Just the Destination ✨
I truly believe that the process of learning and building is far more impactful and educational than just showcasing a polished final product. We all encounter bugs, face imposter syndrome, and hit roadblocks. By being transparent about these parts of the journey, I hope to create a space where it's okay to learn openly and even fail publicly, because that's where the real growth happens.

So, if you're also embarking on your tech journey, are a seasoned pro with wisdom to share, or simply enjoy watching developers build things, I'd be thrilled to connect! Let's learn, build, and grow together.

👉 Follow me here on Dev.to for all my updates, and check out my GitHub Profile to see my code and upcoming projects in real-time.

What is Claude AI — And Why Are IT Companies Panicking?

Siddhartha Kunwar — Wed, 25 Feb 2026 13:12:30 +0000

In the past few months, billions of dollars have been wiped off tech company valuations. Analysts are blaming rising AI capabilities — and one name keeps coming up: Claude AI*. But what exactly is Claude, and why does it matter to YOU — even if you're not a tech person?

🤖 Claude AI in Simple Words

Claude is an AI assistant made by a company called Anthropic. Think of it like a super-smart digital helper — similar to ChatGPT or Google's Gemini — but with some very powerful extra features that are changing how people work.

You can chat with Claude, ask it questions, get it to write emails, create websites, analyse documents, and much more. The best part? You don't need to know any coding to use most of its features.

📉 Why Is It Disrupting the Market?

Claude recently launched features like Skills, Artifacts, and MCP Servers — fancy-sounding terms that basically mean Claude can now do tasks that previously required a whole team of IT professionals.

For example, Claude can:

🌐 Build a complete website from a single sentence
📧 Write and send emails automatically
📁 Read and organise your files
📊 Connect to databases and apps like Google Sheets, Gmail, and Slack

This is why IT stocks dropped — businesses are realising they may need fewer software developers and IT staff when an AI can do the job in seconds.

😮 Should You Be Worried or Excited?

Both!

If you learn how to use Claude well, you become extremely valuable. You can do the work of an entire team by yourself. But if you ignore these tools, you risk falling behind.

The good news is that Claude is designed to be used by everyday people — not just programmers. In this blog series, we'll break down every important feature in plain English so anyone can understand and use it.

🚀 How Do You Get Started?

Simply go to claude.ai and sign up for a free account. Once you're in, you'll see a chat box — just like WhatsApp or iMessage — where you can start talking to Claude right away.

That's it. No downloads, no setup needed for the basics.

✅ Key Takeaway: 🔮 The Bigger Picture

AI tools like Claude won’t replace everyone overnight. But they will change how work gets done.

The real shift isn’t about job loss — it’s about productivity transformation.

The people who learn how to collaborate with AI will outperform those who ignore it.

And that’s where the opportunity lies.

As someone deeply interested in AI and software systems, I believe tools like Claude are not replacing developers — they’re amplifying those who understand them.

This is Blog 1 of a 9-part series breaking down Claude AI for complete beginners. Next up: Claude vs ChatGPT vs Gemini — what actually makes Claude different?