Bart Blast’s Post

Tired of AI coding assistants giving you circular advice? Concerned about your IP leaking through third-party AI tools? Chris Beck's talk from ElixirConf EU 2024 shows how to build your own code intelligence system - entirely in Elixir. 🎯 Key learnings: - How Retrieval Augmented Generation (RAG) works specifically for codebases (it's different from document RAG) - Building RAG systems without relying on Python/JS ecosystems - Maintaining control over your code and data with local tools - Creating a complete Elixir-only RAG system with LiveView https://lnkd.in/dxWQ2xra

How OLSRT breaks languages barriers? Hey Dev Community! Today we want to talk about one of the most exciting aspects of OLSRT: breaking language barriers. Most runtimes are tied to a single language — Node.js for JavaScript, JVM for Java, BEAM for Elixir. This creates silos: developers are forced to choose one ecosystem and stay inside its boundaries. But modern projects are rarely single-language. You might want the speed of C, the safety of Rust, the concurrency of Elixir, or the simplicity of Go — all in one system. That’s where OLSRT comes in. OLSRT is being rebuilt from the ground up to be language-agnostic. Instead of locking developers into one language, it provides a universal runtime that can host multiple ecosystems side by side. For example, OverLab Group is using these languages for OLSRT: C & C++ for low-level control and performance. Rust for memory safety and fearless concurrency. Elixir for distributed, fault-tolerant systems. Go for simplicity and fast networking. ASM for ultimate optimi https://lnkd.in/gJCSJPUJ

TypeScript just became the most-used programming language on GitHub — marking the biggest language shift in over a decade. 🤯 Why? TypeScript helps developers and AI coding assistants catch bugs before code runs, making modern applications safer and faster to build. If you’re building with AI (or plan to!) TypeScript is now essential. Read the full breakdown on what’s behind TypeScript’s rise (and how to learn it): https://bit.ly/4nNV7rJ

Imagine your Java app showing images as they’re being generated — like an AI diffusion model revealing its work in progress. In my latest hands-on tutorial, I walk through building a progressive image streaming service using Quarkus, RESTEasy Reactive, and Mutiny. It’s the same pattern powering modern AI previews and low-latency dashboards. 👉 Read it here: https://lnkd.in/dBNRyXPY #Java #Quarkus #ReactiveProgramming #AI #WebDevelopment #TheMainThread

🤖 Aswanna AI is a chatbot designed to assist users with administrative, educational, technical, and daily inquiries. Built using python backend and responsive frontend provides a sleek and user-friendly interface for real-time communication. Using free API key 🗯️ (⚠️ Token limit applies — reload when the token expires 😊)

[AI News] 🚀 Motia: The Unified Backend Framework Motia brings APIs, background jobs, workflows, and AI agents together in one core primitive—eliminating runtime fragmentation and boosting productivity. Powered by easy multi-language support (JS, TS, Python, Ruby), Motia makes it effortless to build, scale, and connect every backend pattern through Steps—just like React did for frontend components. Key Features: • Unified systems for APIs, queues, jobs, and agents • Native observability & state management • AI development guides and agent integration • Rapid Quickstart and extensible architecture Why it matters: Dev teams can finally stop stitching together a dozen frameworks and start innovating with a single platform and universal primitives. Read more and explore live examples on GitHub. #AIAgents #Backend #Framework #OpenSource #AInews #MasteringLLM

🚀 State of the Champions - Issue #72 is out! This week marks another big step in our mission to make AI agent development more powerful, transparent, and accessible: 🎤 Conf42 JavaScript 2025 I’ve been invited to speak about how JavaScript developers can build intelligent agents that automate workflows and solve real-world problems — all within the ecosystem they already know and love. No ML PhD required. 📦 New in @kaibanjs/opentelemetry package We updated the documentation and added a playground example to make it easy for devs to integrate production-ready observability into multi-agent workflows. ✍️ Deep dives into OpenTelemetry + KaibanJS We shared 3 perspectives on how to make AI agents observable at scale 🔗 Read the full newsletter here → https://shorturl.at/DIy5k #AIagents #JavaScript #OpenTelemetry #Observability #DevTools #Conf42 #KaibanJS

GitHub Octoverse 2025: TypeScript Takes the Lead for the First Time, AI Sets New Development Dynamics TypeScript just overtook Python and JavaScript as the #1 language on GitHub — driven by the AI boom and Copilot adoption. 💡 180M+ developers now use GitHub (1 joins every second). 🤖 200+ new AI projects are created every minute. Read the full Octoverse 2025 report here: https://lnkd.in/dfjaYnwu AI isn’t just helping us code faster - it’s reshaping how we build software. #GitHub #Octoverse2025 #TypeScript #AI #Copilot #SoftwareDevelopment

How To Build WebAssembly Components With the MoonBit Language. A guide to building WebAssembly components with the MoonBit language, from setting up the toolchain to defining interfaces and building the Wasm file.