All posts
Search: “state management” Category: Code and Tech — 8 articles.
Categories
8 articles in Code and Tech · matching “state management”
Code and TechBeyond Demos: The Anatomy of a Robust AI Coding Agent for Long Sessions
Discover how to build a powerful AI coding agent by assembling multiple layers, from orchestration to reflection, to handle complex, real-world development sessions effectively.
Read more
Code and TechReact Compiler as the New Default: What Changes for Developers
The React Compiler is set to become the default for optimizing applications, fundamentally changing how developers approach performance. Learn what this means for your code, from automatic memoization to embracing immutability.
Read more
Code and TechThe Death of "Vibe Coding" and the Rise of Systemic AI Engineering
There is a noticeable shift happening in modern software engineering. For the past year or two, developer culture was dominated by "vibe coding"—a approach where developers prompted AI agents, let LLMs write raw blocks of code, accepted PRs based on visual intuition, and relied on voice-to-text rants to dump context into models. It felt fast, magical, and transformational. However, engineering teams are encountering an inevitable wall: unmaintainable architecture. When AI handles implementation details in isolation, codebases slowly degrade into fragmented, untraceable structures. The defining challenge in modern engineering is no longer how fast can an AI write a function, but how effectively can humans govern the context and architecture that AI agents operate in.
Read more
Code and TechBuilding Agentic Workflows in Python: Beyond Basic RAG with LangGraph
Here is a quick breakdown of what the article covers: * **The Shift:** Traditional RAG follows a rigid, linear pipeline (`Retrieve -> Generate`). **Agentic RAG** introduces dynamic, self-correcting loops using the LLM as a reasoning engine within a state machine. * **Why LangGraph:** Unlike standard DAG (Directed Acyclic Graph) pipelines, LangGraph supports **cyclic workflows** (loops), enabling self-evaluation, error handling, and tool re-invocations. * **Core Architecture:** Built around three elements: * **State:** A shared dictionary (`AgentState`) tracking history and variables across steps. * **Nodes:** Python functions that execute steps (agent reasoning, evaluation, web search). * **Edges:** Conditional routes that direct execution flow based on state evaluation. * **Production Best Practices:** Always enforce **iteration limits** to prevent infinite loops, use **lightweight models** for evaluators to reduce latency, and keep evaluation separate from tool execution.
Read more
Code and TechInfrastructure as Code: Lessons from Our First Terraform Disaster
Learn from our mistakes as we recount a critical incident caused by a misstep with Terraform, highlighting key lessons for robust Infrastructure as Code practices.
Read more
Code and TechThe Edge-First Frontier: How Far Do Edge Functions and Databases Really Go?
Explore the capabilities and limitations of edge functions and databases like Cloudflare Workers and Deno Deploy, and understand when an "edge-first" architecture truly delivers on its promise.
Read more- Code and Tech
Moving Beyond Autocomplete: The Shift from AI Copilots to Multi-Agent Systems (MAS)
The article outlines a major shift in software development: moving from AI Copilots to Multi-Agent Systems (MAS). While traditional copilots simply act as smart autocomplete assistants requiring line-by-line human prompts, Multi-Agent Systems deploy networks of specialized AI agents (e.g., product, coding, testing, and security agents) that work together autonomously to solve complex tasks. Instead of replacing developers, this shift elevates them from manual code-writers to high-level system architects and AI orchestrators.
Read more
