Mom keeping leftovers in the fridge is a cache.
Dinner in 2 minutes. Until nobody checks the date and Tuesday's dal becomes Friday's mistake.
That's a stale cache. TTL is the expiry date nobody reads.
#SystemDesign#Caching
https://t.co/Pb83IegOGC
Your company runs on 5 tools that don't talk: training, shifts, door badges, IT tickets, payroll.
New hire joins: 5 forms. Someone quits: 4 tools still think they work here.
Managemp puts all five on one graph.
#HRTech#HRMS#SaaS
https://t.co/Uwe8GMKK8i
Coming Up Next...
Cuck Meets, where you watch agents join meeting on your behalf to discuss how they can manage your business while you feed them with unlimited tokens for building Apps that only you are going to use ๐ฅฐ
Introducing Claude Meetings: add Claude to your teamโs video call and it builds docs, whiteboards, and other artifacts with you while youโre still talking.
Starting to roll out in beta on Team and Enterprise plans.
Step 95 of 100 fails, and the fix isn't a better prompt.
It's checkpointing, idempotency and retries that don't resend the email.
Agents changed the abstraction layer, not the failure modes.
Brush up on the fundamentals ๐
https://t.co/Pb83IegOGC
#SystemDesign#AIAgents
System design is not important anymore - if you believe this, just wait until your 2-hour agent run fails at step 95 of 100 :) I hope it does not happen, but it actually might.
Agentic apps have two classes: short-running and long-running. When you are building short agents, a simple retry loop is enough. Each LLM call is isolated, cheap, and stateless. Fail, retry, done.
But retrying a long-running agent spanning hours is expensive in API costs, tool calls, and wall clock time, and retries can also have side effects like resending notifications. Some failures that a long-running agent might hit in production are:
- LLM provider rate limiting
- Network timeouts for downstream calls
- Pod eviction due to OOM kills mid-run
- Pod rotation during deployment
- Context window overflow
- Human-in-the-loop pauses for days
Notice how most of these are not AI problems. They are classic software engineering problems that are showing up in a new system.
This is where we need checkpointing, idempotency, durable event logs, and resumable workflows. These are backend engineering fundamentals where we built robust, fault-tolerant systems by applying hundreds of patterns.
Good software engineering practices do not go out of fashion. Because someday, you will ship your "tutorial" to production. That is when system design suddenly matters again.
Fundamentals remain fundamentals, even when the abstraction layer changes.
Yes, DSA is dead.
But a system is not going to build itself yet, is it?
Industry is now #hiring people who can understand the architecture.
Animated system design lessons, a graded HLD playground and 28 interview questions with full answers. No sign-up.
https://t.co/Pb83IegOGC
Electrical Engineers are in the First Phase: DENIAL (where Software Engineers were six months ago)
Coming up Next Phase: "Curiosity and Adaptability" which will later end up with "Acceptance" of AI being better and faster.
Electrical & electronics engineer here:
What Anthropic did is kindergarten style component mash-up not engineering.
Those all are projects a 12 yr old kid can do without even understanding the fundamentals.
If you lead ops, HR or IT at a company with [50โ1,000] employees, Iโd like 15 minutes of your time.
Iโll show you how managemp replaces your HR, payroll, IT desk, access and training tools with one system.
https://t.co/pQu6UrQNZU
DM me โdemoโ ๐
Most coding #interview problems are one of ~20 ideas in disguise.
Interevise teaches the 24 patterns with step-by-step animations, then gives you 168 problems with hidden tests in the browser.
No sign-up. Just start.
https://t.co/xov8pkhFNV
Critical thinking seems to be dwindling as AI adoption progresses.
I feel that a lack of individuality is becoming the norm, as everyone is busy typing prompts all day: โDo this. Make no mistakes.โ ๐ฅ
I want to give you a very interesting insight from seeing video content creators. Most people never succeed at content because theyโre very worried that their friends will find them cringe while they do their first few videos (and get no views).
The secret of the most successful content creators who made a lot of videos early on was that they had no friends and therefore no one to judge. You either needed that, or balls.