Joined June 2020
- Tweets 6,978
- Following 603
- Followers 1,441
- Likes 6,214
650 Photos and videos
Jun 13
Finally back after a long break π
Howβs everyone been? Iβve got some catching up to do
3
32
May 2
Most systems donβt fail because of traffic.
They fail because:
β’ Assumptions break
β’ Edge cases explode
β’ State becomes unmanageable
Scale is not about users.
Itβs about complexity under load.
6
41
K. K. π« retweeted
Apr 6
We just shipped something new on RoastMyProd π₯π¬ A community forum is now LIVEYou can now:
β’ Discuss product ideas
β’ Get real feedback
β’ Share honest roasts
πBuild in public.
Improve faster.
Try it π roastmyprod.com/
4
2
13
261
K. K. π« retweeted
Apr 23
Everything works in dev.
Real engineering starts when:
β’ Traffic spikes
β’ Latency increases
β’ Systems fail
If you havenβt designed for failure β
youβve designed for downtime.
1
4
52
K. K. π« retweeted
Apr 23
Bad architecture is like hidden debt.
You donβt see it early.
But you pay for it with:
β’ Slow features
β’ Fragile systems
β’ Developer frustration
Invest early. Save later.
1
3
53
Apr 21
Everything works in dev.
Real engineering starts when:
β’ Traffic spikes
β’ Latency increases
β’ Systems fail
If you havenβt designed for failure β
youβve designed for downtime.
2
29
Apr 21
The best debugging skill?
Not tools. Not logs.
Itβs the ability to ask:
π βWhat assumptions am I making that might be wrong?β
1
3
33
Apr 20
Your salary doesnβt grow with:
β Hours worked
β Lines of code
It grows with:
β
Impact
β
Ownership
β
Decision quality
Start thinking like a business engineer.
1
29
K. K. π« retweeted
Apr 13
Code reviews are not about catching mistakes.
They are about:
β’ Sharing context
β’ Aligning standards
β’ Preventing future chaos
A good review improves the team, not just the code.
1
2
44
K. K. π« retweeted
Apr 14
Code reviews are not about catching mistakes.
They are about:
β’ Sharing context
β’ Aligning standards
β’ Preventing future chaos
A good review improves the team, not just the code.
1
2
29
K. K. π« retweeted
Apr 14
The best debugging skill?
Not tools. Not logs.
Itβs the ability to ask:
π βWhat assumptions am I making that might be wrong?β
3
5
53
K. K. π« retweeted
Apr 16
I pushed my CAD pipeline to handle 20 STEP files in parallel.
C finished in milliseconds β‘
NestJS crashed in seconds π₯
Lesson:
Fast compute slow transport = system failure.
Fix:
β’ Stream results
β’ Use object storage (MinIO)
β’ Add queue (BullMQ)
Throughput > speed.
1
3
39
Apr 14
Everything works in dev.
Real engineering starts when:
β’ Traffic spikes
β’ Latency increases
β’ Systems fail
If you havenβt designed for failure β
youβve designed for downtime.
21