Writing

Dev.to AI

DeepSeek and the Power of Mixture of Experts (MoE)

DeepSeek is causing a stir in the AI community with its open-source large language models (LLMs), and...

6 min read Read

Dev.to AI

Manus AI: A Technical Deep Dive into China's First Autonomous AI Agent

Manus AI is a pioneering autonomous AI agent developed by Monica, a Chinese AI startup. Unlike...

9 min read Read

Dev.to Mobile

Vibe Coding Flutter: The Senior Dev's Honest Take

I'm a Google Developer Expert in Dart & Flutter. This is my honest, research-backed take on vibe...

11 min read Read

Dev.to AI

Code Knowledge Graphs: How to Evaluate Graphify, GitNexus, and CodeGraph

An evaluation guide for technical leads: the license gate, the staleness window, edge provenance, and what to actually test on your own repo before committing.

5 min read Read

Dev.to AI

Open Knowledge Format (OKF): A Technical Leader's Adoption Guide

Google's Open Knowledge Format packages organizational knowledge as markdown for AI agents. What OKF solves, what it does not, and how to adopt it.

5 min read Read

Dev.to Mobile

Flutter Material and Cupertino Leave the SDK: What Changes for Your App

Material and Cupertino are becoming material_ui and cupertino_ui on pub.dev. Here is the real timeline, the pubspec change, and what to do now.

5 min read Read

Medium Engineering Craft

You Can Outsource the Team. You Cannot Outsource the Consequences.

Read

Dev.to AI

Agent Design Patterns: Google and Anthropic, Side by Side

A field guide to agent design patterns from Anthropic and Google, with when to use each, the trade-offs, and animated diagrams.

5 min read Read

Dev.to AI

Autonomous Agents and the ReAct Loop: When the Model Owns Control

Autonomous agents run a think, act, observe loop, Google's ReAct pattern. When to hand control to the model, and how to keep it safe.

4 min read Read

Dev.to AI

Composite Patterns: How Real Agent Systems Combine the Basics

Composite and custom logic patterns combine routing, parallelization, and critic loops into one system. When to mix patterns and how to keep control.

4 min read Read

Dev.to AI

Human-in-the-Loop: Let a Person Authorize the Risky Step

The human-in-the-loop pattern pauses an agent at a checkpoint so a person can approve high-stakes actions. When to use it and what it costs to build.

4 min read Read

Dev.to AI

The Swarm Pattern: Peer Agents That Debate and Converge

The swarm pattern uses all-to-all agent collaboration with no central boss. Google's most flexible and most costly pattern, with its real failure modes.

4 min read Read

Dev.to AI

Evaluator-Optimizer: Generate, Critique, and Refine in a Loop

The evaluator-optimizer, Google's generator-critic and iterative refinement loop, improves output through feedback. When it helps and its failure modes.

4 min read Read

Dev.to AI

Orchestrator-Workers: Decide the Subtasks at Runtime

Orchestrator-workers, Google's hierarchical task decomposition, lets a lead agent split a task at runtime and delegate. When to use it and its costs.

4 min read Read

Dev.to AI

Parallelization: Run Subtasks at Once, Then Merge

Parallelization, Google's fan-out and gather pattern, runs independent subtasks at the same time and merges them. Sectioning, voting, and the trade-offs.

4 min read Read

Dev.to AI

Routing: Classify the Input, Then Send It to a Specialist

Routing, Google's coordinator pattern, classifies each input and sends it to a specialized agent. When it beats one big prompt, and its trade-offs.

4 min read Read

Dev.to AI

Prompt Chaining: Fixed Steps That Each Feed the Next

Prompt chaining, Google's sequential pipeline pattern, breaks a task into fixed steps with a gate between them. When to use it and where it breaks.

4 min read Read

Dev.to AI

The Augmented LLM: The Building Block Every Agent Sits On

The augmented LLM is the base unit of every agent: one model plus retrieval, tools, and memory. When to use it, and when a single agent is enough.

5 min read Read

Dev.to AI

The New SDLC for Project Managers: Vibe Coding vs Agentic Engineering

A project manager's guide to the AI-driven SDLC: how vibe coding and agentic engineering change estimates, cost, risk, and how teams are staffed.

5 min read Read

Dev.to Mobile

Dart FFI for Flutter: Call Native C Directly (Lessons from camera_pro)

Dart FFI lets Flutter call native C code directly, without platform channels. A short, practical guide with diagrams, plus real lessons from building the camera_pro package.

5 min read Read

Dev.to AI

The LLM Preamble Problem: How RLHF Made Your Model Too Polite to Ship

Instruction-tuned LLMs add 'Certainly!' preambles before every answer. Here is the RLHF root cause, the real cost in tokens, and four developer fixes.

11 min read Read

Dev.to AI

GGUF, Explained: The File Format That Put LLMs on Your Laptop

GGUF is the format that lets large language models run on a laptop, a phone, even a Raspberry Pi. What it is, what the quant names mean, and how to choose.

14 min read Read

Dev.to AI

AI Gateways: A Senior Engineer's Honest Take

An AI gateway sits between your apps and LLM providers. What it does, how it differs from an API gateway, and the core features actually worth paying for.

16 min read Read

Dev.to Mobile

Flutter GenUI SDK: Build AI-Generated UIs from Scratch (Complete Beginner Tutorial 2026)

Step-by-step beginner tutorial: build real-time AI-generated Flutter widget trees using the GenUI SDK, Gemini, and custom CatalogItems. No prior AI experience needed.

19 min read Read

Dev.to Mobile

I Built Kalaam: An Arabic Tutor Whose UI Gemini Assembles at Runtime

Kalaam is an open-source Flutter app where Gemini composes each Arabic lesson interface at runtime from a catalog of 27 widgets. Clone, run in 30 seconds, and watch AI-native UI work.

18 min read Read

Dev.to AI

Spring AI: The Senior Dev's Honest Take on Java's AI Moment

Spring AI connects enterprise Java to AI models with portable abstractions, RAG, MCP, and tool calling. Here is what it actually does and when to reach for it.

13 min read Read

Dev.to AI

Is Token Usage the New Lines of Code? How to Measure Developer Productivity in the AI Age

Token budgets are becoming the new productivity flex in Silicon Valley. But is measuring token usage just as broken as counting lines of code? Here is what the research and the sharpest minds in engineering actually say.

7 min read Read

Dev.to Mobile

The New SDLC: A Senior Dev's Honest Take on Vibe Coding and Agentic Engineering

The SDLC didn't get faster, it got different. How agentic engineering, context engineering, and the factory model reshape software delivery in 2026.

15 min read Read

Dev.to AI

LLM Token Counting Explained: The 30-Year-Old Algorithm Quietly Running Your AI Bill

TLDR: A token is a subword chunk, typically 3 to 4 characters in English. The algorithm splitting...

5 min read Read

Dev.to Mobile

Antigravity 2.0 for Flutter Developers: CLI, SDK & Agentic Workflows That Actually Matter

Google's Antigravity 2.0 ships a new CLI, SDK, and Flutter MCP server at I/O 2026. Here's the practical playbook for Flutter devs to ship faster with AI agents.

5 min read Read

Dev.to AI

I Pointed a Skill Linter at a 52k-Star Repo. Here Is What 84/100 Looks Like.

I ran skillscore on all 24 skills from addyosmani/agent-skills to find what separates a 91/A from a 77/C. Two patterns showed up in every single C. Both are fixable in under 10 lines.

5 min read Read

Dev.to AI

DiffusionGemma: How Google's New Open LLM Hits 1,000 Tokens/sec and Changes Inference Economics

DiffusionGemma generates text up to 4x faster than autoregressive LLMs, hits 1,000+ tokens/sec on a single H100, and runs on a consumer RTX 4090. Here is what changed, what the trade-offs are, and how to deploy it today.

4 min read Read

Dev.to Mobile

Flutter Agent Skills: How to Make Your AI Agent Actually Good at Flutter

TL;DR: Your AI coding assistant is a generalist. It writes Flutter that looks right but quietly...

11 min read Read

Dev.to Mobile

How to Write a Flutter Agent Skill That Actually Works: The 2026 Recipe

TL;DR A great agent skill is not a pile of documentation. It is a tightly scoped SKILL.md with a...

14 min read Read

Dev.to AI

skillscore: a CLI that scores your AI agent's SKILL.md 0-100

skillscore is an open-source Dart CLI that lints and scores agent SKILL.md files against the Claude, Codex, and Antigravity authoring guides , offline, deterministic, CI-ready.

6 min read Read