Collaborative Trial: On Optimizing Recommendation Testing
Taboola is responsible for billions of daily recommendations, and we are doing everything we can ...
Data & Analytics
59 Articles
How Taboola Handles Deep Learning Training-Serving Features Discrepancies
Introduction In machine learning, one critical challenge is ensuring that the features used during model ...
LLM-Gateway: Streamlining Large Language Model Integration at Taboola – Part 1
Introduction & Overview As the landscape of LLMs rapidly evolves, leading providers such as OpenAI, ...
Zooming Past the Competition
Imagine you’re walking down the street and you see a nice car you’re thinking of ...
Driving Success Through Cross-Functional Management in Engineering at Taboola – Part 2: Strategic Leadership and Innovation in Tracks Methodology
Introduction In the first part of this series, we explored how the Tracks methodology empowers ...
Driving Success in Engineering at Taboola: Cross-Functional Management Blueprint
In today’s fast-paced business environment, flexibility and collaboration are key to driving innovation and delivering ...
Optimizing Ad Space Through Mixing Organic Items In The Open Web
Introduction Taboola is the world’s leading native advertising platform, serving thousands of websites with a ...
Optimizing CTR Prediction through Clustering-Based Feature Engineering
With more than 500,000 recommendations per second, Taboola is the world’s leading native advertising platform ...
Utilizing LLMs for Localized Recommendations
Do you ever go to a local boutique brewery and buy a Carlsberg? Drive to ...
Real-time Language Sync: Web App to Browser Extension
In previous articles, I went into an efficient implementation of internationalization using react-i18next for browser ...
Breaking the Scale Barrier – Smarter Test Selection
A build process is a critical procedure triggered by new code changes (Git push). Its ...
GPU Integration Propels Data Center Efficiency and Cost Savings for Taboola
A year ago, we at Taboola kicked off a POC to migrate our Spark workload ...