Shashi Jakka । శశి జక్క @ssJvirtually

Antigravity CLI Useful Commands Cheat Sheet.

DSA @neetcode1 System Design @bytebytego Backend @hnasr Frontend @frontendmasters only resources one needs. Rest is practice, practice, and practice

A must read article for Spring Boot application developers by @vlad_mihalcea Spring Boot Application Properties vladmihalcea.com/spring-boot-ap…

quitting asp!!

Generally "reinventing the wheel" is not a recommended practice while building production systems. But, for learning anything in-depth "reinventing the wheel" is a good approach. We can simply use Claude Code CLI and ask any questions about the repository, generate README, etc. I thought it could be a good idea to implement the same using Spring AI to explore some core AI concepts deeply. So, I tried doing RAG by ingesting all the source code into a vector database and doing Q&A. Then I got to know that Tree Sitter based parsing is better suited than simple size-based chunking. github.com/sivaprasadredd… Its working okayish :-) At Spring I/O Mark Pollack gave a workshop on automating the PR review process using Spring AI github.com/markpollack/ag… So much to explore and learn... Reinventing the wheel is good for learning.

Building agent-ready RAG in Java without pretending enterprise PDFs are already nice Markdown. #Quarkus + #Docling + pgvector + #Ollama, with background indexing, readiness, retrieval logging, and guardrails. Small enough to run locally, shaped like something you can extend. buff.ly/JpI936E

So freaking cool. I have Claude Code, Cowork & Travis (OpenClaw) all working together, talking to eachother, giving eachother tasks, using Obsidian as a shared brain. Creating more openclaw subagents next. I couldn't even imagine this stuff two years ago.

A Rust dev just killed Headless Chrome. It's called Obscura. The open-source headless browser purpose-built for AI agents and scrapers at scale. Chrome vs Obscura: - Memory: 200MB+ → 30MB - Binary: 300MB+ → 70MB - Page load: 500ms → 85ms - Startup: 2s → Instant - Anti-detect: None → Built-in Single binary. No Node, no Chrome, no dependencies. Stealth mode is brutal: → Per-session fingerprint randomization (GPU, canvas, audio, battery) → 3,520 tracker domains blocked by default → navigator.webdriver masked to match real Chrome → Native function masking so detectors can't sniff it out Drop-in replacement for Puppeteer and Playwright over CDP. Zero code changes. If you run agents or serious scraping at scale, this repo prints money. 100% Opensource.

Karpathy didn't make a course. He made THE course. 3 hours. Free. Tokenization. Attention. Hallucinations. Tool use. RLHF. DeepSeek. AlphaGo. Every behavior you've ever wondered about in an LLM - where it comes from, why it exists, how it was engineered. The gap between engineers who understand this and engineers who don't isn't technical depth. It's the ability to conceive of entirely different things.

93.5% of Senior Backend Engineer interviews are just these 7 concepts repeated again and again.

PDFs are officially dead. Someone just built a tool that turns PDFs into clean, structured Markdown at 100 pages/sec 🤯 No GPU. No API cost. No messy parsing. Just raw, usable data. Here’s what it handles effortlessly: • Tables → perfectly extracted • Broken layouts → auto-fixed • Nested data → structured cleanly • Scanned chaos → turned readable This isn’t a small upgrade. This wipes out 90% of manual data cleaning overnight. The tool is OpenDataLoader And yes… it’s open-source. Repo → github.com/opendataloader…

Two-Phase Commit (2PC) Achieving Strong Consistency Across Services The Core Problem In microservices, each service has its own database. Example: Order Service DB Payment Service DB Inventory Service DB Now consider a transaction: Place Order → Deduct Payment → Reserve Inventory What if: - Payment succeeds - Inventory fails System becomes inconsistent. What 2PC Solves Two-Phase Commit ensures: Either all services commit OR all services rollback No partial success allowed. 2PC Architecture There are two roles: 1. Coordinator Central controller of the transaction. Responsible for: - Asking participants to prepare - Deciding commit/rollback 2. Participants Services involved in transaction. Example: Payment Service Inventory Service Order Service Two Phases of 2PC 1. Phase 1 - Prepare Phase (Voting) Coordinator asks: Can you commit? Each participant: - Executes transaction locally - Locks resources - Responds: YES (ready) NO (fail) 2. Phase 2 - Commit Phase If ALL say YES: Coordinator → COMMIT If ANY say NO: Coordinator → ROLLBACK Example Flow Step-by-step: 1. Coordinator → Payment: Prepare? 2. Coordinator → Inventory: Prepare? 3. All say YES 4. Coordinator → Commit all If failure: Inventory says NO → rollback everything Problems with 2PC 1. Blocking Problem If coordinator crashes: - Participants wait indefinitely. - System gets stuck. 2. Performance Issues - Requires locking resources - Slow (two network round trips) - Not scalable 3. Not Cloud-Friendly Modern distributed systems avoid 2PC because: - High latency - Low availability - Poor fault tolerance 2PC vs Modern Systems Feature 2PC Consistency Strong Availability Low Performance Slow Scalability Limited Where 2PC Is Used Still used in: - Banking systems - Legacy enterprise systems - Distributed SQL databases Architect-Level Insight 2PC guarantees: Strong consistency But sacrifices: Availability + performance This is why modern systems prefer: - Saga Pattern - Eventual consistency Final Insight 2PC is theoretically correct but practically restrictive. Modern systems move from: Strict consistency to Eventual consistency with compensation

Being a good #Java developer is not the same as being a strong interview candidate. If you’re stuck between interviews and offers, you don’t need more grinding. You need clarity. Book a Career Diagnosis Session here: bit.ly/4j5lLfb?utm_so…

keyboard_arrow_left Previous keyboard_arrow_right Next

🔥 Spring Boot looks simple… until the interviewer goes deep. Here are **real Spring Boot internals questions** asked in recent interviews 👇 1. How does Spring Boot decide which **auto-configuration** to apply? 2. What happens internally when you add `spring-boot-starter-web`? 3. Why does Spring Boot prefer **convention over configuration**? 4. How does Spring Boot load `application[dot]properties` internally? 5. Exact **startup flow** of a Spring Boot application. 6. Difference between `@ComponentScan` and `@SpringBootApplication`. 7. How does Spring Boot detect **embedded Tomcat** and configure it? 8. What happens if two beans of the same type exist without `@Qualifier`? 9. How does Spring Boot handle **profile-specific configuration**? 10. What is the role of `SpringFactoriesLoader` under the hood? 11. Difference between `@RestController` and `@Controller` internally. 12. How does Spring Boot manage **dependency versions automatically**? 13. Lifecycle of a Spring Bean in Spring Boot. 14. How does Spring Boot handle **externalized configuration**? 15. Fat jar vs normal jar — internal difference. 16. How Spring Boot decides **server port priority**. 17. What happens internally when you hit a **REST endpoint**. 18. How Spring Boot integrates with **Actuator** internally. 19. How exception translation works in Spring Boot. 20. Common performance mistakes in Spring Boot applications. 💡 **Pro Tip:** Knowing **internals** is what separates average from strong candidates. If you can explain 10–15 of these confidently, you’re already ahead of 90% of Spring Boot developers. 📌 Save this for interviews

high performance, instant startup Gemma4 inference in pure Java & @GraalVM!🔥

Devoxx @Devoxx

2 months ago

🚀 Gemma4 Java runner: • Single file, no deps • E2B→31B + MoE • GGUF + quant (F16→Q8) • Vector API ⚡ • CLI + thinking modes • GraalVM native + instant TTFT Pure Java. No excuses 😏 github.com/mukel/gemma4.j… by @TheMukel ☕️🔥

1 16 48 7K 38

Good to see a couple of articles I have written are referenced in @testcontainers skill reference materials: github.com/spring-ai-comm…

My repo (👉github.com/piomin/claude-…) with Claude Code template for Spring Boot with instructions, skills, and subagents just crossed 1k⭐️

Gemma 4 can run on phones without an internet connection! 🤯 It can perform local agentic tasks, such as logging and analyzing trends. When connected, it can also make API calls. Want to try it yourself? Get the Google AI Edge App on iOS or Android. (🔊 Sound on for the demo!)

8 RAG architectures for AI Engineers: (explained with usage) 1) Naive RAG - Retrieves documents purely based on vector similarity between the query embedding and stored embeddings. - Works best for simple, fact-based queries where direct semantic matching suffices. 2) Multimodal RAG - Handles multiple data types (text, images, audio, etc.) by embedding and retrieving across modalities. - Ideal for cross-modal retrieval tasks like answering a text query with both text and image context. 3) HyDE (Hypothetical Document Embeddings) - Queries are not semantically similar to documents. - This technique generates a hypothetical answer document from the query before retrieval. - Uses this generated document’s embedding to find more relevant real documents. 4) Corrective RAG - Validates retrieved results by comparing them against trusted sources (e.g., web search). - Ensures up-to-date and accurate information, filtering or correcting retrieved content before passing to the LLM. 5) Graph RAG - Converts retrieved content into a knowledge graph to capture relationships and entities. - Enhances reasoning by providing structured context alongside raw text to the LLM. 6) Hybrid RAG - Combines dense vector retrieval with graph-based retrieval in a single pipeline. - Useful when the task requires both unstructured text and structured relational data for richer answers. 7) Adaptive RAG - Dynamically decides if a query requires a simple direct retrieval or a multi-step reasoning chain. - Breaks complex queries into smaller sub-queries for better coverage and accuracy. 8) Agentic RAG - Uses AI agents with planning, reasoning (ReAct, CoT), and memory to orchestrate retrieval from multiple sources. - Best suited for complex workflows that require tool use, external APIs, or combining multiple RAG techniques. 👉 Over to you: Which RAG architecture do you use the most? _____ Share this with your network if you found this insightful ♻️ Find me → @akshay_pachaar ✔️ For more insights and tutorials on LLMs, AI Agents, and Machine Learning!

someone built a web-based System Design Simulator. you drag and drop components (api gateways, dbs, caches) and it actually simulates real-time traffic. you can watch latency, bottlenecks, and failures happen live...