SDE @Microsoft | Ex-Infosys | 133K+ on LinkedIn | 2000+ rated on LeetCode | Code, Content and Memes

I'm a Software Development Engineer at Microsoft and an ex-Infosys employee. I have over 159k followers on LinkedIn and I'm passionate about coding, creating content, and sharing memes. I'm passionate about helping others learn and grow in their professional career.


Codechef 5 star ( 2003 ) | Hackerrank 6 star | HackerEarth 1918 | Leetcode 2112 | CodeForces specialist | Highest rated coder in college | SDE at  Microsoft | Ex-Infosys | Techgig-Code Gladiator finalist 2019 , 2020 ,2021, 2022, 2023,2024

Bhen ke bhaiyo ladna band krdo 😭😭😭

#market #war

Chandan Agrawal

Many people are criticizing the candidate in this story, but I see it differently.

Hiring is a two-way street.

Companies often take days or weeks to respond, pause hiring, or move on without updates. Candidates are expected to simply wait and understand.

But when a candidate keeps exploring better opportunities, suddenly it becomes unethical.

Also, companies themselves frequently base offers on current salary. Two people can join the same role and do the same work, yet one might earn 2x more just because of their previous pay.

Let’s be honest - money runs life. People naturally try to secure the best opportunity they can.

In this market we’ve also seen lowball offers, delayed joining (especially in campus hiring), and even offer revocations.

So if companies optimize for their interests, it’s fair that candidates do the same.

The real answer is simple: clear communication and mutual respect on both sides.

Money is important specially for men, got rejected ( many of my friends too) for being at low package 🥲

#hiring #job #hiring #viral

Jeeeeeet gaye ❤️😇

 #IndVsNz #t20worldcup

No alternative text description for this image

𝐅𝐢𝐫𝐬𝐭 𝐈 𝐰𝐢𝐥𝐥 𝐬𝐞𝐞 𝐂++, 𝐭𝐡𝐞𝐧 𝐈 𝐰𝐢𝐥𝐥 𝐬𝐞𝐞 𝐲𝐨𝐮 ❤️❤️

She was clearly giving signals…
But I was busy understanding Advanced C++.

And honestly, 𝐂++ bhi kuch kam toxic nahi hai 💀

Here are some “reasons” why 👇

1️⃣ 𝐏𝐨𝐢𝐧𝐭𝐞𝐫𝐬 & 𝐑𝐞𝐟𝐞𝐫𝐞𝐧𝐜𝐞𝐬
Relationship bhi pointer jaisa hi hota hai…
Galat handle kiya -> Segmentation fault 💥

2️⃣ 𝐌𝐞𝐦𝐨𝐫𝐲 𝐌𝐚𝐧𝐚𝐠𝐞𝐦𝐞𝐧𝐭
Manual ho ya relationship…
Free karna bhool jao → Memory leak 😶

3️⃣ 𝐔𝐧𝐝𝐞𝐟𝐢𝐧𝐞𝐝 𝐁𝐞𝐡𝐚𝐯𝐢𝐨𝐫
C++ me bhi unpredictable…
Aur relationships me bhi 🙂

4️⃣ 𝐓𝐞𝐦𝐩𝐥𝐚𝐭𝐞𝐬 (𝐆𝐞𝐧𝐞𝐫𝐢𝐜𝐬)
“Ek hi code sab pe kaam karega”
Reality: Edge cases ruin everything

5️⃣ 𝐌𝐮𝐥𝐭𝐢𝐭𝐡𝐫𝐞𝐚𝐝𝐢𝐧𝐠 & 𝐑𝐚𝐜𝐞 𝐂𝐨𝐧𝐝𝐢𝐭𝐢𝐨𝐧𝐬
Tum soch rahe ho kuch aur…
Wo expect kar rahi hai kuch aur…
→ Deadlock 🤝

6️⃣ 𝐂𝐨𝐧𝐬𝐭 𝐂𝐨𝐫𝐫𝐞𝐜𝐭𝐧𝐞𝐬𝐬
Log bolte hain “I’ll never change”
C++ bolta hai: const ka matlab samajh pehle 😌

7️⃣ 𝐓𝐢𝐦𝐞 𝐂𝐨𝐦𝐩𝐥𝐞𝐱𝐢𝐭𝐲
Simple lagta hai…
But worst case -> O(n²) emotions 😭

Moral of the story:
Debugging C++ is still easier than decoding signals 😌

#meme #tech #dsa #cpp #coding #dev #faang #microsoft

Finally, after relentless research and development
at my personal Centre of Excellence…
.
.
.
.
I have successfully achieved
50 Continuous Days of LeetCode. 🏅💻

𝐏𝐫𝐚𝐜𝐭𝐢𝐜𝐞 𝐃𝐒𝐀 𝐥𝐢𝐤𝐞 𝐚 𝐩𝐫𝐨 : www.algotracker.in

#meme #tech #ai #coding #leetcode

Netflix migrated 150+ million records under 40K RPS 😲😲

𝐍𝐞𝐭𝐟𝐥𝐢𝐱 executed one of the cleanest large scale database migrations in production.
‣ 150+ million records
‣ 40K requests per second
‣ 300 GB of live data
‣ Zero downtime

𝐓𝐡𝐞 𝐅𝐞𝐚𝐭𝐮𝐫𝐞: 𝐍𝐞𝐭𝐟𝐥𝐢𝐱 𝐐𝐮𝐞𝐮𝐞 🎬
The Queue was a core user feature.
It allowed members to:
• Maintain an ordered list of movies and shows
• Store activity metadata
• Power personalization and recommendations
Slow or inconsistent data directly impacted user experience.

⚠️ 𝐖𝐡𝐲 𝐌𝐢𝐠𝐫𝐚𝐭𝐞 𝐟𝐫𝐨𝐦 𝐒𝐢𝐦𝐩𝐥𝐞𝐃𝐁?
𝐒𝐢𝐦𝐩𝐥𝐞𝐃𝐁 worked well initially, but scale changed everything.
Challenges emerged:
• Latency spikes
• SLA limitations
• Throttling under heavy load
• Rising operational cost
• Not ideal for write heavy workloads

𝐍𝐞𝐭𝐟𝐥𝐢𝐱 needed:
• High write throughput
• Low latency
• Better scalability
• Predictable performance

𝐂𝐚𝐬𝐬𝐚𝐧𝐝𝐫𝐚 became the target datastore.
Constraint: No downtime

😬 𝐖𝐡𝐲 𝐌𝐢𝐠𝐫𝐚𝐭𝐢𝐨𝐧 𝐢𝐬 𝐇𝐚𝐫𝐝
Live systems never pause.
Users keep reading
Users keep writing
Data keeps changing
Copy once → New DB becomes stale
Switch abruptly → Risk of outages & data loss

𝐍𝐞𝐭𝐟𝐥𝐢𝐱 avoided the big bang approach.

🚚 𝐒𝐭𝐞𝐩 𝟏: 𝐅𝐨𝐫𝐤𝐥𝐢𝐟𝐭 𝐌𝐢𝐠𝐫𝐚𝐭𝐢𝐨𝐧
Full snapshot copy:
SimpleDB → 𝐂𝐚𝐬𝐬𝐚𝐧𝐝𝐫𝐚
Moved historical data, but live updates continued.

🔁 𝐒𝐭𝐞𝐩 𝟐: 𝐈𝐧𝐜𝐫𝐞𝐦𝐞𝐧𝐭𝐚𝐥 𝐑𝐞𝐩𝐥𝐢𝐜𝐚𝐭𝐢𝐨𝐧
Only changed records synced using:
Last_Updated_TS
Cassandra gradually caught up with live mutations.

🔍 𝐒𝐭𝐞𝐩 𝟑: 𝐂𝐨𝐧𝐬𝐢𝐬𝐭𝐞𝐧𝐜𝐲 𝐂𝐡𝐞𝐜𝐤𝐞𝐫
Failures were expected.
They continuously compared:
SimpleDB vs Cassandra
Mismatch detected → Repaired

✍️ 𝐒𝐭𝐞𝐩 𝟒: 𝐒𝐡𝐚𝐝𝐨𝐰 𝐖𝐫𝐢𝐭𝐞𝐬
Writes → SimpleDB + Cassandra
Reads → Still SimpleDB
Safely tested Cassandra under real traffic.

👀 𝐒𝐭𝐞𝐩 𝟓: 𝐒𝐡𝐚𝐝𝐨𝐰 𝐑𝐞𝐚𝐝𝐬
Reads → SimpleDB + Cassandra
Responses compared asynchronously.
Mismatch metrics tracked.

🔄 𝐅𝐢𝐧𝐚𝐥 𝐂𝐮𝐭𝐨𝐯𝐞𝐫
When mismatches < 0.01 percent:
Reads → Cassandra
Writes → Cassandra
SimpleDB retired.
No drama
No outage
No user impact

🧠 𝐊𝐞𝐲 𝐋𝐞𝐬𝐬𝐨𝐧𝐬
Migration is not copy paste
Design for failure
Use forklift + incremental sync
Shadow traffic reduces risk
Validate with metrics before switching

💡 𝐅𝐢𝐧𝐚𝐥 𝐓𝐡𝐨𝐮𝐠𝐡𝐭
Good engineering chooses new tech.
Great engineering migrates without breaking reality.

If you enjoy real world scaling stories, follow Chandan for more 🚀

#design #migration #netflix #blog #systemdesign #coding #meme #tech #ai #google #microsoft #uber #salesforce

𝐌𝐨𝐬𝐭 𝐜𝐨𝐝𝐢𝐧𝐠 𝐢𝐧𝐭𝐞𝐫𝐯𝐢𝐞𝐰𝐬 𝐛𝐞 𝐥𝐢𝐤𝐞 - 

𝗜𝗻𝘁𝗲𝗿𝘃𝗶𝗲𝘄𝗲𝗿:“Solve Two Sum.”
𝗠𝗲 : opens HashMap confidently 😌

𝐓𝐰𝐨 𝐒𝐮𝐦 is that classic friend who just wants a pair.
𝐓𝐡𝐫𝐞𝐞 𝐒𝐮𝐦 is the dramatic upgrade. Now we need a trio.

Quick recap:
--------------

✅ 𝐓𝐰𝐨 𝐒𝐮𝐦
 Store numbers in a map
 Check target minus current
 Done in O(N)

✅ 𝐓𝐡𝐫𝐞𝐞 𝐒𝐮𝐦
 Sort the array
 Fix one number
 Use two pointers
 Welcome to O(N²)

Same idea. More chaos. More loops. 😄

Funny how one extra number turns “easy win” into “think harder”.

Follow Chandan for more 😄😄

#meme #tech #dsa #ai #leetcode

Two sum and Three Sum

Most people practice LeetCode in a way that feels productive… but doesn’t actually prepare them for interviews.

A few habits that make practice much more effective:

1️⃣ 𝐃𝐨𝐧’𝐭 𝐣𝐮𝐬𝐭 𝐬𝐨𝐥𝐯𝐞 - 𝐮𝐧𝐝𝐞𝐫𝐬𝐭𝐚𝐧𝐝 𝐭𝐡𝐞 𝐩𝐚𝐭𝐭𝐞𝐫𝐧
Sliding window, monotonic stack, prefix sum, two pointers, BFS/DFS… many questions are just variations of the same ideas. Recognizing patterns is what saves time in interviews.

2️⃣ 𝐒𝐨𝐥𝐯𝐞 𝐰𝐢𝐭𝐡 𝐚 𝐭𝐢𝐦𝐞𝐫
Give yourself 25 – 40 minutes per problem. Interviews are timed, and practicing under time pressure trains your decision making.

3️⃣ 𝐔𝐬𝐞 𝐠𝐨𝐨𝐝 𝐯𝐚𝐫𝐢𝐚𝐛𝐥𝐞 𝐧𝐚𝐦𝐞𝐬 𝐚𝐧𝐝 𝐰𝐫𝐢𝐭𝐞 𝐜𝐥𝐞𝐚𝐧 𝐜𝐨𝐝𝐞
Interviewers look at readability. Clear variable names and structured code make your thought process easier to follow.

4️⃣ 𝐄𝐯𝐞𝐧 𝐢𝐟 𝐲𝐨𝐮 𝐬𝐨𝐥𝐯𝐞𝐝 𝐢𝐭, 𝐫𝐞𝐚𝐝 𝐨𝐭𝐡𝐞𝐫 𝐬𝐨𝐥𝐮𝐭𝐢𝐨𝐧𝐬
Sometimes there’s a cleaner trick, a different data structure, or a more optimal approach you didn’t think about.

5️⃣ 𝐓𝐫𝐲 𝐯𝐢𝐫𝐭𝐮𝐚𝐥 𝐜𝐨𝐧𝐭𝐞𝐬𝐭𝐬 𝐛𝐞𝐟𝐨𝐫𝐞 𝐫𝐞𝐚𝐥 𝐨𝐧𝐞𝐬
Do ~20 virtual contests with a strict 1-hour timer. It builds speed, pressure handling, and prioritization.

6️⃣ 𝐏𝐫𝐚𝐜𝐭𝐢𝐜𝐞 𝐦𝐨𝐜𝐤 𝐢𝐧𝐭𝐞𝐫𝐯𝐢𝐞𝐰𝐬
You can do them with friends, platforms like pramp.com, or paid sessions on Topmate.

7️⃣ 𝐑𝐞𝐜𝐨𝐫𝐝 𝐲𝐨𝐮𝐫𝐬𝐞𝐥𝐟 𝐞𝐱𝐩𝐥𝐚𝐢𝐧𝐢𝐧𝐠 𝐭𝐡𝐞 𝐬𝐨𝐥𝐮𝐭𝐢𝐨𝐧
Try recording your voice or video while solving. You’ll quickly notice where you hesitate, overthink, or struggle to explain clearly.

LeetCode is not about solving the most problems.

It’s about training how you think, communicate, and perform under pressure.

Follow Chandan for more 😊😊

Yes, Noida office me MS logo ke saath pic nahi lene dete to...........

#leetcode #codinginterview #softwareengineering #datastructures #algorithms #tech #dsa #google #maang #coding

𝐖𝐡𝐲 𝐜𝐨𝐩𝐲𝐢𝐧𝐠 𝐚 𝐡𝐮𝐠𝐞 𝐟𝐢𝐥𝐞 𝐚𝐠𝐚𝐢𝐧 𝐚𝐧𝐝 𝐚𝐠𝐚𝐢𝐧 𝐢𝐬 𝐚 𝐛𝐢𝐠 𝐰𝐚𝐬𝐭𝐞📦💸
.
.
.
Imagine this situation.

You have a 10 GB log file.

Every 2 hours your system downloads and copies it to another server.

But most of the time the file does not change.
Maybe only 100 KB of new logs were added.

Yet the system still transfers the full 10 GB again.

𝗧𝗵𝗮𝘁 𝗺𝗲𝗮𝗻𝘀 𝗮 𝗹𝗼𝘁 𝗼𝗳 𝘄𝗮𝘀𝘁𝗲𝗱:
❌ Network bandwidth
❌ Disk I/O
❌ Time

This is where 𝐫𝐬𝐲𝐧𝐜 becomes very powerful.
💡 The smart idea behind 𝐫𝐬𝐲𝐧𝐜

Instead of copying the whole file, 𝐫𝐬𝐲𝐧𝐜 only 𝘁𝗿𝗮𝗻𝘀𝗳𝗲𝗿𝘀 𝘁𝗵𝗲 𝗰𝗵𝗮𝗻𝗴𝗲𝗱 𝗽𝗮𝗿𝘁𝘀 of the file.

So if only a small portion changed, only that part moves across the network.

Think of it like this.

𝐍𝐨𝐫𝐦𝐚𝐥 𝐜𝐨𝐩𝐲 works like:
📦 Send the 𝗲𝗻𝘁𝗶𝗿𝗲 𝗯𝗼𝘅 𝗮𝗴𝗮𝗶𝗻 𝗲𝘃𝗲𝗻 𝗶𝗳 𝗼𝗻𝗹𝘆 𝗼𝗻𝗲 𝗶𝘁𝗲𝗺 𝗶𝗻𝘀𝗶𝗱𝗲 𝗰𝗵𝗮𝗻𝗴𝗲𝗱.

𝐫𝐬𝐲𝐧𝐜 works like:
📦 Send 𝗼𝗻𝗹𝘆 𝘁𝗵𝗲 𝗶𝘁𝗲𝗺 𝘁𝗵𝗮𝘁 𝗰𝗵𝗮𝗻𝗴𝗲𝗱.

📊 Simple Example
Suppose this file exists on two servers.

𝐎𝐥𝐝 𝐟𝐢𝐥𝐞 𝐨𝐧 𝐝𝐞𝐬𝐭𝐢𝐧𝐚𝐭𝐢𝐨𝐧
[A][B][C][D][E]

𝐍𝐞𝐰 𝐟𝐢𝐥𝐞 𝐨𝐧 𝐬𝐨𝐮𝐫𝐜𝐞
[A][B][𝗫][D][E]

Only C changed to X.

𝐍𝐨𝐫𝐦𝐚𝐥 𝐜𝐨𝐩𝐲
----------------
System transfers : A B X D E
Full file transfer.

𝐑𝐬𝐲𝐧𝐜
-------
System detects that A B D E already exist.
So it transfers only - 𝐗

Destination rebuilds the final file using existing blocks.
𝐑𝐞𝐬𝐮𝐥𝐭
[A][B][𝐗][D][E]

Much faster and much less data moved.

📉 Real world numbers

Imagine this scenario.
File size
 📦 20 GB
Change in file
 ✏️ 5 MB
Runs per day
 ⏰ 12 times

Using 𝐍𝐨𝐫𝐦𝐚𝐥 𝐜𝐨𝐩𝐲
- 𝟐𝟎 𝐆𝐁 × 𝟏𝟐 = 𝟐𝟒𝟎 𝐆𝐁 𝐭𝐫𝐚𝐧𝐬𝐟𝐞𝐫 𝐩𝐞𝐫 𝐝𝐚𝐲

Using 𝐫𝐬𝐲𝐧𝐜
- 𝟓 𝐌𝐁 × 𝟏𝟐 = 𝟔𝟎 𝐌𝐁 𝐭𝐫𝐚𝐧𝐬𝐟𝐞𝐫 𝐩𝐞𝐫 𝐝𝐚𝐲

That is a massive improvement.

🧠 How 𝐫𝐬𝐲𝐧𝐜 does this
𝐑𝐬𝐲𝐧𝐜 breaks files into small blocks.

Then it compares those blocks between the source and destination.
If a block already exists on the destination, 𝐫𝐬𝐲𝐧𝐜 reuses it.
If a block changed, 𝐫𝐬𝐲𝐧𝐜 transfers only that block.
So the final file is rebuilt using a mix of
 🔹 existing blocks
 🔹 new blocks
This technique is called delta transfer.

🌍 𝐖𝐡𝐞𝐫𝐞 𝐭𝐡𝐢𝐬 𝐢𝐝𝐞𝐚 𝐢𝐬 𝐮𝐬𝐞𝐝
This concept powers many modern systems.
 📁 Backup systems
 ☁️ Cloud storage sync
 📦 Linux package mirrors
 📂 Dropbox style file sync
 🐙 Git style data transfer

All of them try to transfer only what actually changed.
⚡ One simple takeaway

Before moving huge data, ask one question. Did anything actually change?

If the answer is small, only move the difference.

That is the magic of 𝐫𝐬𝐲𝐧𝐜.

Follow Chandan for more such in depth and interesting things .

#tech #dsa #systemdesign #design #meme #fun #google #microsoft

taplio

Chandan Agrawal's Linkedin Analytics

Chandan Agrawal

Performance ScoreChandan Agrawal's LinkedIn Performance

Chandan Agrawal's Best Posts (last 30 days)

Ready to grow your LinkedIn presence?

taplio

.css-6ytdur{background-image:linear-gradient(284deg, rgba(39, 43, 72, 1) 0%, rgba(33, 47, 183, 1) 100%);color:transparent;-webkit-background-clip:text;background-clip:text;}Chandan Agrawal's Linkedin Analytics

Chandan Agrawal

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Chandan Agrawal's LinkedIn Performance

Chandan Agrawal's Best Posts .css-1emoo84{font-size:var(--chakra-fontSizes-md);font-weight:400;color:var(--chakra-colors-gray-600);}(last 30 days)

Ready to grow your LinkedIn presence?

Chandan Agrawal's Linkedin Analytics

Performance ScoreChandan Agrawal's LinkedIn Performance

Chandan Agrawal's Best Posts (last 30 days)