Tag
A guide to 40 backend concepts organized into six stages, from HTTP basics to system architecture, designed to help engineers prepare for system design interviews.
A pull request adds an initial ET backend to llama.cpp, expanding hardware support for LLM inference.
User seeks advice on running llama.cpp with Gemma 4 E2B on an Intel N100 mini PC, asking whether to use CPU or iGPU and which backend to target.
Vercel introduces Service Bindings, allowing private connectivity between backend services using internal URLs configured in vercel.json with a single line of code.
Encore announces embedding a Redis server directly into their runtime, simplifying development by eliminating the need for an external Redis instance.
A detailed guide on building a production-grade agent harness for multi-agent LLM systems, covering components like orchestrator, subagents, skills, backend state management, and context engineering.
Introduces performance improvements by reducing synchronizations during split compute in the ggml backend, adds async CUDA copy capability, and makes sync relaxation more general across backends.
A list of 10 backend development projects to build for learning, including notification systems, video processing queues, and search autocomplete.
Zig's SPIR-V backend has been overhauled with new built-in types, improved execution mode handling, capabilities/extensions from CPU features, multi-threaded codegen, and object file linking, making it more useful for shader and GPU programming.
The author reflects on their mistaken belief that AI agent cost is solely a backend problem, suggesting a broader perspective on cost factors.
A high-performance instant messaging backend based on Rust + Actix-web + QUIC, supporting P2P hole punching, group chat broadcasting, multiple storage backends, can be deployed as a single process or separately.
BlitzGraph is a backend service for graph databases, designed as an AI-native alternative to Supabase tailored for LLM agents.
A developer claims to have built a full backend (Google OAuth, Stripe payments, AI integration, database, deployment) in 2 hours using AI, offering a breakdown via DM.
The blog post details why the team chose Elixir over Go, Rust, or Node for building a performant Bluesky DataPlane, leveraging the BEAM's concurrency for the hot path and offloading compute-intensive operations to a Rust NIF.
Claude Fable 5, likely the public-facing version of Claude Mythos 5, has been spotted on Azure and the backend.
A Twitter thread listing 35 essential system design concepts with links to detailed explanations, aimed at helping developers learn and review key topics.
A detailed technical deep-dive into the Rust-powered backend of Nosdesk, covering architecture decisions like streaming pipelines, Postgres sync, and type-safe design patterns across 120k lines of code.
The article discusses how AI products require a new 'AI integration layer' to handle context retrieval, tool execution, model routing, and observability, and references Merge.dev's infrastructure for this purpose.
QBE 1.3 is a significant compiler backend release with 7k new lines of code, featuring a new IL matching algorithm, optimizations for coremark benchmark (improving from 40% to over 63% of gcc -O2 performance), Windows ABI support, and position-independent code generation.
A technical walkthrough of solving a concurrent device registration race condition without adding new infrastructure like Redis, using database-level locking and creative schema handling.