dan72ros
128.5K posts

dan72ros
@dan72ros
ideas, politics, religion and sometimes dark humor connected to life in Israel. Also, enjoys: steak, the beach https://t.co/delyZIvth5
Israel Katılım Ekim 2008
1.7K Takip Edilen1.8K Takipçiler
dan72ros retweetledi

8 types of LLMs used in AI agents 🤖
GPT • MoE • LRM • VLM • SLM • LAM • HRM • LCM
Different models for reasoning, perception, planning, and action — not just chat.
Agentic AI = model orchestration.
#AI #LLMs #AgenticAI #GenAI #MachineLearning

English
dan72ros retweetledi

Want to become a Claude Certified Architect in just 6 weeks?
Here’s a no-BS roadmap to go from learning → building → certified
Week 1 Foundations
→ Learn the core:
• Claude API
• Model Context Protocol (MCP)
• Claude Code
• Claude 101
Week 2 Build
→ Stop consuming. Start creating:
• Real projects with Claude Code
• Agent SDK + APIs
• MCP integrations
Week 3 Understand the Exam
→ Know the game before you play:
• 6 real-world scenarios
• 5 core domains
• Skills needed for each
Week 4 Practice Like a Pro
→ Build what actually matters:
• Multi-tool AI agents
• Team workflows with Claude Code
• Data pipelines + research agents
Week 5 Test Yourself
→ Take mock exams
→ Target 850+/1000
Week 6 Go Time
→ Attempt the real exam
→ One shot. Make it count.

English
dan72ros retweetledi

🧵 Day 7/30
Your app works with 10 servers.
What happens when you scale to 1,000 servers?
How do you distribute users/data evenly without chaos?
That’s where Consistent Hashing comes in.
What is Consistent Hashing?
It’s a technique used to distribute data across multiple servers efficiently.
Instead of assigning data randomly, both:
→ Servers → Keys (users, cache items, requests)
are placed on a logical hash ring.
Each key goes to nearest server.

Nitin.nn@NitinthisSide_
🧵 Day 6/30 — #SystemDesign Your database works fine… until users grow 100x. Then queries slow down, writes queue up, and one machine becomes the bottleneck. That’s where Database Sharding comes in. ⸻ What is Sharding? Sharding means splitting one large database into multiple smaller databases called shards. Instead of storing all users in one DB: → Shard 1 stores some users → Shard 2 stores others → Shard 3 stores others Data is distributed horizontally.
English
dan72ros retweetledi
dan72ros retweetledi
dan72ros retweetledi

I understand the logic of punishing soldiers who had a barbeque on Shabbat. This arrest? Simple idiocy.
timesofisrael.com/idf-court-mart…
English

I wish I understood the allure of this pro wrestling stuff. But ESPN? The sports leader? I know I’m gonna get blasted as soon as I hit send but I don’t get it. What are we doing? Where are we going?
Awful Announcing@awfulannouncing
ESPN’s embrace of WWE has officially gone overboard awfulannouncing.com/espn/embrace-w…
English
dan72ros retweetledi

AI Agents = LLMs + orchestration 🚀
Here are the main types of LLMs powering them 👇
🧠 General-purpose (GPT, Claude)
⚙️ Domain-specific (Legal, Finance, etc.)
📚 RAG-based (real-time knowledge)
🤖 Tool-augmented (API actions)
🧩 Open-source (LLaMA, Mistral)
The game is no longer the model.
It’s how you combine them. 🔥
#AI #LLM #AIAgents #GenAI

English
dan72ros retweetledi

System Design Roadmap
│
├── 1. Introduction to System Design
│ ├── What system design is
│ ├── Importance in real-world applications
│ └── Key goals: scalability, reliability, performance
│
├── 2. Fundamental Concepts
│ ├── Networking basics (HTTP, HTTPS, DNS)
│ ├── Client-server architecture
│ └── Latency vs throughput
│
├── 3. Databases & Storage
│ ├── SQL vs NoSQL
│ ├── Data modeling
│ ├── Indexing
│ └── Replication and sharding
│
├── 4. Scalability Principles
│ ├── Vertical scaling
│ ├── Horizontal scaling
│ └── Load distribution
│
├── 5. System Components
│ ├── Load balancers
│ ├── Application servers
│ ├── Databases
│ ├── Cache (Redis, Memcached)
│ └── CDN
│
├── 6. Caching Strategies
│ ├── Client-side caching
│ ├── Server-side caching
│ └── Cache invalidation
│
├── 7. Distributed Systems
│ ├── CAP theorem
│ ├── Consistency models
│ ├── Fault tolerance
│ └── Consensus basics
│
├── 8. API Design
│ ├── RESTful APIs
│ ├── Rate limiting
│ └── Authentication & authorization
│
├── 9. Message Queues & Async Systems
│ ├── Event-driven architecture
│ ├── Queues (Kafka, RabbitMQ)
│ └── Background processing
│
├── 10. Designing Real Systems
│ ├── URL shortener
│ ├── Chat system
│ ├── File storage system
│ └── News feed system
│
├── 11. Monitoring & Reliability
│ ├── Logging
│ ├── Metrics and alerts
│ └── High availability
│
└── 12. Interview Strategy
├── Clarify requirements
├── High-level design first
├── Deep dive into components
└── Discuss trade-offs
→Grab the System Design Roadmap Handbook
→ codewithdhanian.gumroad.com/l/urcjee

English

So Israel stole land…
Well, I have some good news for you.
We will gladly give the land back! No, seriously.
But only after YOU the world give your land back to the native people you stole it from.
When the Australians give the island back to the Aborigines.
And the New Zealanders give it back to the Maori.
When the Americans and the Canadians and the Spanish give the whole continent back to the Natives.
Alaska to the Eskimos.
Hawaii to Hawaiians.
When China frees Tibet.
When South Africa returns to the original Africans.
When Egypt return Sinai to the Bedouin and Pakistan will be India again.
When the Swedes and Norwegians and Swedes give the land to the Sami.
When the English frees Ireland and Scotland.
When France returns to the Gauls
and Iraq to Babylonians.
And Persia is Sumer again
Then and only then will we gladly give our land back.
But not to the watermelon people 🍉
but to those who were there before us.
To the Canaanites, the Amorites, the Moabites, the Girgashites, the Hittites, the Midianites and the Levites.
but…
Alas!
They no longer exist.
So we will stay here.
But YOU can happily go back to where ever YOU came from.
And if you don't like it, well… Keep your mouth shut.
- The Jewish people who returned to their land after thousands of years.
Chag Sameach and Am Israel Chai!
HT @BearShai for putting this together.

English
dan72ros retweetledi

Open more than the average Israeli post office 🤷🏼♂️
Brandon Beylo@marketplunger1
lol who did this.
English










