Blog — React, Next.js & TypeScript articles by Exequiel Sosa
// Notes on Front-end engineering
Thoughts and explorations on React, Next.js, TypeScript, performance and the tools I use every day. New post every week.
Why gpt-5.6-luna’s 100‑token‑per‑second speed matters for front‑end engineers
Discover how the cheap, fast gpt-5.6-luna model reshapes AI‑augmented UI, on‑device inference, and cost‑effective RAG pipelines for modern front‑end development.
[nextjs]Decoupling Marketing Copy with the ExperienceSlot Pattern in Next.js
Learn how the ExperienceSlot pattern separates copy from React components, cuts bundle bloat, and eliminates hydration bugs in large Next.js apps.
[ai]What OpenAI’s Zero Data Retention Means for Front‑End Engineers
Explore how OpenAI’s new zero data retention policy affects API usage, security, and UX design for React and Next.js developers, with practical tips and code snippets.
[react]Why TanStack Table v9 Beta fixes INP lag by tree‑shaking unused grid features
Discover how TanStack Table v9’s modular API lets you ship only the table features you need, cutting bundle size and eliminating Interaction‑to‑Next‑Paint stalls in production React apps.
[ai]How Prompt Caching Cuts LLM Costs for Front‑End Teams
Learn practical prompt caching tricks to shrink token usage and keep your AI‑augmented front‑end pipelines under budget, with real code examples.
[react]Common React Performance Mistakes You’re Probably Making in Code Reviews
A no‑fluff guide to the most frequent React performance pitfalls seen in code reviews, with practical fixes and real‑world examples.
[performance]Mastering CSS writing-mode in React: practical patterns for vertical text layouts
Learn how to use CSS writing-mode in React and Next.js projects. Real‑world examples, performance tips, and trade‑offs for vertical text in modern web apps.
[ai]How Finite State Machines Keep Your AI Publishing Agents Reliable
Discover why adding a finite state machine to AI agents prevents missed steps, loops, and interruptions—using a real dev.to publishing bot as a practical example.
[react]Understanding React Flight: Risks, Deserialization Sinks, and Safe Patterns
Dive into the React Flight protocol behind Server Components, discover deserialization pitfalls, and learn practical defenses for production apps.
[ai]Testing Non-Deterministic AI UIs: Practical Evals for Front‑End Teams
Learn how front‑end engineers can reliably test AI‑driven, non‑deterministic interfaces with evaluation pipelines, synthetic data, and real‑user feedback loops.
[nextjs]Eliminate Data Waterfalls in Next.js with Parallel Fetching and Suspense
Learn how to break data waterfalls in Next.js App Router using parallel fetches, Server Components, and React Suspense for faster, smoother user experiences.
[ai]LLM Gateway Showdown: LiteLLM vs OpenRouter vs Portkey for Front‑end Teams
Compare LiteLLM, OpenRouter, and Portkey as LLM gateways. Learn trade‑offs, deployment models, and how each impacts your React/Next.js stack and cost monitoring.
[react]Optimistic UI with React useOptimistic: Patterns That Scale in 2026
Learn practical ways to use React 19's useOptimistic hook for instant feedback, error rollback, and real‑world performance trade‑offs in production apps.
[ai]What GPT-5.6 and Full‑Duplex Voice Mean for Front‑End Engineers
Explore how the new GPT-5.6 models and GPT‑Live full‑duplex voice impact front‑end development, from cost‑effective APIs to real‑time voice UI patterns.
[nextjs]Real Costs of Migrating WordPress to Next.js in 2026 – What to Expect
A practical look at the price ranges for moving a WordPress site to Next.js in 2026, with trade‑offs, hidden costs, and real‑world migration patterns for teams.
[nextjs]Dynamic Page Builder with Strapi, Next.js, and Varnish: Real-World Trade-Offs
Learn how to let editors assemble pages in Strapi while keeping Next.js fast and cache‑friendly with Varnish. Practical patterns, pitfalls, and performance tricks for production.
[ai]Why Your Front‑End LLM App Needs an Evaluation Layer (and How to Build One)
Learn the practical evaluation layer every production LLM UI misses, with real front‑end patterns, code snippets, and trade‑offs for reliable AI responses.
[ai]Choosing an EU‑Hosted LLM Inference Provider in 2026: Practical Front‑End Guide
A front‑end engineer’s rundown of the top EU‑hosted inference services, pricing, latency, and GDPR trade‑offs to help you pick the right provider for production apps.
[react]Mastering React Suspense and the use() Hook
Unlock the full potential of React Suspense and the use() hook for seamless async UI management and data fetching
[ai]Designing AI Systems: A New Era of Interview Questions
Get ready for AI system design interviews with ChatGPT, RAG, and LLM inference. Learn how to approach these new challenges with practical examples and insights.
[performance]May Web Platform Updates
Explore the latest web platform features that landed in stable and beta browsers in May 2026, and what they mean for you as a developer
[ai]Mastering Prompt Engineering for AI Developers
Discover the top concepts for effective prompt engineering to unlock reliable AI outputs, including task analysis and context engineering
[frontend-news]Streamlining Accessibility Labels
Improving screen reader experience with concise navigation labels and alt text
[ai]Unlocking AI-Powered Coding with North Mini Code
Discover how North Mini Code, Cohere's first model for developers, is changing the game for front-end engineers and explore its potential applications
[performance]Fixing Color Contrast Issues with CSS
Improve web accessibility by leveraging the contrast-color() function to build self-correcting color systems, reducing WCAG contrast failures
[ai]Practical AI in Healthcare: Improving Length-of-Stay Predictions
Discover how ChatHealthAI achieves 79.8% F1 on length-of-stay prediction using a frozen LLM and task-aware resampler, and what this means for front-end developers
[performance]The Hidden Cost of Optimizations
Uncovering the unintended consequences of front-end optimizations, and how a small change can lead to significant performance issues
[ai]Image Generation with Flux API
Generate images using Flux API in Python with a simple tutorial and examples
[react]Asynchronous Requests in Event-Driven Frontends
Decoupling components with event buses for asynchronous requests and responses in real-world applications
[react]Demystifying Roku's ECP API
Building an in-browser Roku TV remote with TypeScript and exploring the External Control Protocol
[react]Under the Hood of useLocalStorageState
Dive into the internals of the useLocalStorageState hook and explore how it combines useState and localStorage for persistent state management
[ai]Testing Non-Deterministic UIs with AI Features
Learn how to evaluate and test AI-powered UI components in your front-end applications, ensuring reliability and user experience.
[react]Server vs Client Components: Weighing Your Options
Understand the trade-offs between server and client components in modern web development to make informed decisions for your project
[ai]Vercel AI SDK: Lessons from Production
Real-world experiences and lessons learned from integrating Vercel AI SDK into production applications, improving user experience and streamlining development
[performance]Beyond Band-Aids: Fixing Core Web Vitals
Discover why your Core Web Vitals remain subpar despite applying obvious fixes and learn how to tackle deeper issues for significant improvements
[frontend-news]Elevating Chatbot UX
Discover how to create a chatbot UI that feels premium and engaging, exploring key UX patterns and front-end techniques for a seamless user experience
[typescript]TypeScript Generics In Practice
Explore the practical use of TypeScript generics in real-world applications, including their benefits and potential drawbacks
