🍔🧠 1 Embedding Model That Powers Airbnb's Entire Search System

Happy Monday! ☀️

Welcome to the 464 new hungry minds who have joined us since last Monday!

If you aren't subscribed yet, join smart, curious, and hungry folks by subscribing here.

📚 Software Engineering Articles

• The free guide to manage dev operations

• Learn how to prevent system misuse and protect resources effectively

• Breaking down circular dependencies for cleaner JavaScript code

• Remove jQuery from legacy applications step by step

• Why one developer quit using AI code editors

• Customize select elements with pure CSS, finally!

🗞️ Tech and AI Trends

• Apple's AI doctor could revolutionize personal healthcare

• Amazon bids to acquire TikTok's US operations

• Gemini 2.5 sets new performance records in AI

👨🏻‍💻 Coding Tip

• Use RedisJSON instead of string serialization for efficient JSON storage

Time-to-digest: 5 minutes

Big thanks to our partners for keeping this newsletter free.

If you have a second, clicking the ad below helps us a ton—and who knows, you might find something you love.

💚

Optimize global IT operations with our World at Work Guide

Explore this ready-to-go guide to support your IT operations in 130+ countries. Discover how:

• Standardizing global IT operations enhances efficiency and reduces overhead

• Ensuring compliance with local IT legislation to safeguard your operations

• Integrating Deel IT with EOR, global payroll, and contractor management optimizes your tech stack

Leverage Deel IT to manage your global operations with ease.

Download free guide

1 Embedding Model That Powers Airbnb's Entire Search System 🏠️ 

Airbnb built a powerful search system that leverages machine learning to match millions of homes with the right guests. Their embedding-based retrieval (EBR) approach maps both search queries and listings into a shared mathematical space, enabling fast and accurate matches at a massive scale.

The challenge: Create a retrieval system that can efficiently narrow down millions of listings while capturing complex search intent and maintaining sub-second latency.

Implementation highlights:

• Smart training data: Built using actual user booking journeys, capturing both booked homes and intentionally rejected options

• Two-tower architecture: Separate networks for query and listing features, with listing embeddings pre-computed daily

• Efficient serving: Used IVF clustering for fast nearest neighbor search compatible with filters

• Euclidean distance: Chose over dot product for better cluster balance

• Contrastive learning: Trained model to maximize similarity between queries and booked homes vs non-booked options

Results and learnings:

• Significant impact: Achieved one of the largest booking gains from an ML improvement in the past 2 years

• Query-aware: System effectively incorporates search context for better relevance

• Production-ready: Successfully deployed in both Search and Email Marketing systems

The success of Airbnb's EBR system shows that sophisticated ML techniques can be practical at scale when carefully designed. Like a great matchmaker, it helps connect millions of guests with their perfect stays, proving that sometimes the best technology is the one you don't even notice working.

ESSENTIAL (cursor-controlled chaos)
What Are Cursor Rules?

GITHUB REPO (minecraft-server-madness)
awesome-mcp-servers

GITHUB REPO (node-version-ninja)
nvm

ARTICLE (hooks-but-fancy)
Components Are Just Sparkling Hooks

ARTICLE (next-js-oopsie)
Postmortem on Next.js Middleware bypass

ARTICLE (vibe-check-failed)
Karpathy's ‘Vibe Coding' Movement Considered Harmful

ARTICLE (select-no-more-ugly)
The select element can now be customized with CSS

ARTICLE (next-js-houdini)
Deploying a Next.js App to Production in any server

ARTICLE (jquery-yeet)
Mission jQuery Zero: How FreeAgent removed jQuery from our application

ARTICLE (sync-stopper-3000)
Stop syncing everything

ARTICLE (ai-no-thanks)
Why I stopped using AI code editors

ARTICLE (circular-javascript-drama)
Breaking down circular dependencies in JavaScript

ARTICLE (anchor-link-overkill)
Overengineered anchor links

Want to reach 170,000+ engineers?

Let’s work together! Whether it’s your product, service, or event, we’d love to help you connect with this awesome community.

👯‍♂️ Meta Revives 'OG Facebook' Vibes with New Friends Tab (2 min)

Brief: Meta launches a Friends Tab to prioritize close connections over algorithm-driven content, as Zuckerberg aims to recapture Facebook's early social-focused roots with more nostalgic updates planned.

🤖 Elon Musk Merges xAI With X in $33B Deal to Unite AI and Social Media (2 min)

Brief: Musk’s xAI acquires X (formerly Twitter) in a $33B all-stock deal, combining AI tech with social media’s reach while raising concerns over data integration and regulatory scrutiny.

🤯 Gemini 2.5 Pro Dethrones Rivals as New AI Benchmark King (4 min)

Brief: Google's Gemini 2.5 Pro emerges as the new state-of-the-art LLM for reasoning and coding, outperforming rivals like Claude 3.7 and GPT-4o in benchmarks but being criticized for censorship and poor documentation.

👓 Meta's $1,000 Smart Glasses With a Screen: How Will They Work? (2 min)

Brief: Meta's next-gen $1,000 smart glasses will reportedly feature a built-in display, advancing the company’s push into wearable AR tech despite stiff competition.

🛒 Amazon Joins the Race to Acquire TikTok Amid Rising Geopolitical Tensions (2 min)

Brief: Amazon enters the bidding war for TikTok as the US pressures ByteDance to divest its stake, potentially reshaping the social media and e-commerce landscape.

This week’s coding challenge:

This week’s tip:

Store Redis data in a more memory-efficient way using RedisJSON instead of storing serialized JSON strings. RedisJSON provides native JSON support with efficient memory usage and allows for direct JSON path operations without parsing.

Wen?

• Large JSON documents: When dealing with complex nested JSON structures that need frequent partial updates or reads.

• Memory optimization: When memory usage is critical and you're storing many JSON documents that are frequently accessed partially.

• Performance-critical operations: When you need fast access to nested JSON properties without the overhead of serialization/deserialization.

We are what we repeatedly do. Excellence, then, is not an act but a habit. Aristotle

That’s it for today! ☀️

Enjoyed this issue? Send it to your friends here to sign up, or share it on Twitter!

If you want to submit a section to the newsletter or tell us what you think about today’s issue, reply to this email or DM me on Twitter! 🐦

Thanks for spending part of your Monday morning with Hungry Minds.
See you in a week — Alex.

Icons by Icons8.

*I may earn a commission if you get a subscription through the links marked with “aff.” (at no extra cost to you).