Back to blog
Reading path· 8 articles

AI Integration Path

A frontend-first guide to adding AI features without becoming an accidental ML platform team.

Start with the AI pattern overview, then go deeper into RAG and streaming interfaces.

1
Tier 2 16 min read

5 AI Patterns Every Frontend Engineer Will Build in 2026

A frontend-first guide to the five AI product patterns that are becoming standard, and how to decide which one fits your app.

2
Tier 2 17 min read

The Frontend Engineer's Guide to RAG — No ML Degree Required

A frontend-first introduction to retrieval-augmented generation, with practical architecture choices and evaluation ideas.

3
Tier 2 18 min read

Building a Streaming AI Chat UI — The Complete Frontend Architecture Guide

A production-minded guide to streaming AI chat interfaces covering transport, buffering, cancellation, error recovery, and trust cues.

4
Tier 1 16 min read

Client-Side ML Inference — Running AI Models in the Browser Without Melting the UI

A practical guide to running TensorFlow.js and Transformers.js models in the browser — model loading strategies, Web Worker inference, quantization trade-offs, and device-aware degradation.

5
Tier 1 15 min read

Embeddings and Vector Search in the Browser — A Frontend Engineer's Guide

Understand embeddings from first principles, then build vector search entirely in the browser using Transformers.js, cosine similarity, IndexedDB, and Chrome AI APIs.

6
Tier 1 17 min read

Tool-Calling Agents and the ReAct Pattern — Building AI That Takes Action

How tool-calling agents work, the ReAct reasoning loop, when to use agents vs chains, and how to design approval gates and rollback for production agent systems.

7
Tier 2 14 min read

AI Content Moderation for Frontend Engineers — Pipelines, Classification, and Safety UX

How to design a multi-stage content moderation system that combines client-side pre-screening, server classification, and human review with transparent, trustworthy UX.

8
Tier 2 14 min read

Structured AI Output and Intelligent Forms — Getting Clean Data from LLMs

How to use structured output, schema-constrained generation, and extraction pipelines to turn unstructured text into typed form data with confidence scoring.