#recommendation-system

Who’s your neighbour?

How we rebuilt end-of-story recommendations at Pratilipi — 1.3 million series, 12 languages, and one three-hour outage.

GenRec: Towards LLM-Native Recommendation at Netflix

Authors: Ying Li, Arjun Rao, Shradha Sehgal

Tavsiye Sistemlerinden A/B Testine: Association Rule Learning, Content-Based ve Collaborative…

Birliktelik kuralları, Apriori algoritması, içerik temelli filtreleme, iş birlikçi filtreleme ve deneysel karar verme süreçleri

McDonald’s’tan Yediğimiz Her Şey Sağlıksız Mı? Merakımdan Başlayan Küçük Bir Veri Analizi

Geçtiğimiz günlerde evde otururken kendime iki McDonald’s menüsü söyledim. Yemeğin ardından aklıma tek bir soru takıldı: “Acaba şu an…

Designing a Three-Tier Cold-Start Strategy with Fold-In Approximation and Online Blending

How I built an on-device book recommendation engine that gracefully handles the transition from “I know nothing about you” to “I know…

Sepetten Öneriye: Python ile Association Rule Learning

Python ile Association Rule Learning: Armut Verisi Üzerinden Hizmet Tavsiye Sistemi

Measuring the Impact of Personalized Recommendations

By Kevin Zielnicki, Guy Aridor, Aurélien Bibaut, Allen Tran, Winston Chou, and Nathan Kallus

Hibrit Öneri Sistemi: MovieLens 20M Verisiyle Adım Adım User-Based + Item-Based Collaborative…

Bu yazıda, 20 milyon satırlık MovieLens 20M veri setini kullanarak sıfırdan bir hibrit öneri sistemi (hybrid recommender system) kurma…

Movie Recommender using Cosine Similarity

A recommendation system is a software that predicts what items a user might like and suggests them accordingly. Example Netflix suggestions…

What Makes Songs Popular on Spotify? I Analyzed 114,000 Songs to Find Out

Every day, Spotify recommends songs that somehow seem to match our taste. Some tracks become global hits, while others receive very little…

Building a Multimodal Vector Search Platform for Product Catalogs (and What Bit Us Along the Way)

What broke, what surprised us, and what we'd keep: multitenancy, GPU economics, and the boring parts nobody blogs about.

Learning Recommendation Systems by Building a Hybrid Movie Recommender

How I built Movie Recommendation system to explore content-based filtering, collaborative filtering, and the machine learning concepts…

ML Deep Dive: Multitask Learning

Multitask learning is all the rage these days, being especially useful and widely adopted in modern ranking systems like ads, search, and…

GenPage: Towards End-to-End Generative Homepage Construction at Netflix

Authors: Lequn Wang, Jiangwei Pan, and Linas Baltrunas

Real-Time Personalisation at Scale: How Zepto Understands What You Want, Right Now

Zepto started as a quick-commerce platform for groceries. Today it’s where people shop for electronics, home essentials, gifts and festive…

Cómo construir un sistema de recomendación híbrido

En el primer artículo de esta serie vimos que un sistema de recomendación no es simplemente un modelo de Machine Learning, sino una…

Cómo construir un sistema de recomendación Content-Based

En el primer artículo de esta serie vimos que un sistema de recomendación no es simplemente un modelo de Machine Learning, sino una…

Waveform-Gated Recommendations — Why Your Festival Content Is Failing (and the Fix Is Not More…

This article is inspired by the challenges encountered while designing and implementing a Template Intelligence Engine (also referred to as…

Beyond Cosine Similarity: Building an Adaptive Hybrid Recommendation Engine

When I started building CineAI, my first instinct — like most people’s — was to throw together a content-based recommender using cosine…

Pinterest’s Recommender System Is Becoming a Foundation Model: A Deep Dive into PinRec and PinFM

How Pinterest is rethinking recommendation systems with generative retrieval, user-sequence foundation models, outcome conditioning, and…

Ontology Meets Graph Databases: Semantic Personalization at Scale for Recommendation Systems

Ontology types and weights relationships, embedding semantic meaning into graphs.

No 2. Beyond the “Lookalike” Trap: The Hidden Bottleneck in LLM-Driven Recommendations

We’ve all seen it: “Users who bought X also bought Y.” It’s the classic, familiar lookalike recommendation pattern that has powered…

Welcome to Machine Learning With Manya: The Ultimate Adventure Map!

If you have ever stared at a mathematical formula or a machine learning textbook and thought, “There has got to be a better way to…

Your Company Intranet Is Where Knowledge Goes to Die. Here’s the Engine I’d Build to Fix It.

A practical architecture for an enterprise recommendation system — LightGBM, a hybrid batch + real-time pipeline, and a “why this?” button…

Limited Memory AI-The Second Level

The Intelligence behind ChatGPT, Netflix & Self-Driving Cars. This level shows how AI learns from past data to make smarter decisions.

PRS 2026: What the Industry Learned About Personalization, Recommendation & Search in the LLM Era

Notes from the 10th annual Netflix PRS workshop, where DoorDash, LinkedIn, Pinterest, Google DeepMind, Stanford, Zillow, and Netflix shared…

The Algorithm Loves You and Wants to Help You

The first time Marcus noticed it was mentioning a knee. He’d said it out loud in the kitchen, the way you say things when you’ve spent lots…

That 4.9 Star Product Is Lying to You

I am sure by the end of this article you will know exactly why you choose a 4.4 star product with 10,000 reviews over a 4.9 star product…

Counterfactual Evaluation in Ads: IPS, SNIPS, and Doubly Robust

You have a new ranking model. You want to know if it’s better than the one in production before you ship it. The honest answer is: you…

What Protein Folding Taught Me About User Modeling

A structural biologist’s framework for thinking about recommendation systems

Bias Analysis Report: Iran-US Conflict Coverage on TikTok (US Data)

Figure 1a — Pro-US group bias vs. Monte Carlo null distribution

The Quiet Cost of Letting AI Decide

Most discussions about AI revolve around what it can do.

LeetCode taught us to solve, LeetPal will help with what to solve next!

Try Out LeetPal!

I Built a Recommendation Engine in 30 Minutes (And So Can You)

A system that recommends things by comparing users with similar interest are everywhere in social media.

The Computer Science Behind Recommendation Systems

Every day, recommendation systems quietly shape our decisions.

YouTube Isn’t Reading Your Mind — Here’s How It Predicts Your Next Video

Have you ever opened YouTube intending to watch just one video, only to find yourself still scrolling an hour later?

Learning-to-Rank (Chapter 1): Ranking Fundamentals

Let us look at two real-world examples where ranking models are used.

E-Commerce Recommendation System: End-to-End Case Study

In this article, we will see on how to approach case study round in the top product companies in the interviews for the position of data…

Deep semantic embeddings

Deep semantic embeddings are vector representations that capture the meaning of an entity — such as a user, an ad, a query, or a product —…

Collaborative Filtering: Wisdom of the Crowd, Personalised for You

Imagine walking into a bookstore and asking the staff: “I loved Dune — what should I read next?” The best recommendations don’t come from…

Content-Based Filtering: How Machines Learn Your Taste

Ever noticed how Spotify keeps suggesting songs that sound similar to what you just played? Or how Netflix recommends films with the same…

How Recommendation Systems Learn From What You Ignore

Your Silence Is Also Data

System Design for Beginners #20: How Recommendation Systems Work (How Apps Predict What You Want)

Imagine opening YouTube.

My Impact in Building Pulse-Agent (An AI Agent That Reads Your Mind — DSN x BCT Hackathon 3.0)

Zero Followers. 4 Days. 10 Million People. Here’s the Engineering That Made It Possible.

10 million followers. 4 days. Zero followers to start. This isn’t a content story — it’s an engineering story.

I Built a Product Recommendation System for an FMCG Company. Here’s Where It Started (Part 1)

From a basic collaborative filtering prototype to realizing “technically working” and “actually useful” are two very different things.

The Recommendation System Needed More Than Good Metrics, it Needed to Actually Know Its Customers.

How layering category based personalization on BPR finally made the system feel like it understood who it was talking to.

The Architecture of Virality: How Instagram Builds Million-Follower Empires Overnight

This week, an Instagram account exploded from:

Memory-Driven Learning Assistant Engineering Deep Dive

API, 데이터 모델, 모델 라우팅, 운영/평가 전략까지 엔지니어 시점에서 본 전체 구조

How Spotify Decides What Song to Play Next

Introduction