Subvolts take Meta engineers reveal the complex build behind Facebook Reels' "Friend Bubbles." May 8, 2026 - Developers
- Learn about scaling social features, performance optimization with prefetching, and ML model adaptation for user signals.
- Business
- Understand how iterative product development and user feedback can refine features for better engagement.
- Everyday users
- See how seemingly simple features like friend bubbles are built and improved for your social media experience.
This episode provides a solid interview with engineers detailing the development of a significant consumer-facing feature.
85: Reel Friends: Building Social Discovery that Scales to Billions
Audio streams from Meta. Your place is remembered on this device.
From Subvolts
Summary and analysis by Subvolts editorial, AI-assisted.
Meta Tech Podcast episode 85 features engineers Joseph and Subasree discussing "Friend Bubbles," a Facebook Reels feature. Host Pascal Harris explores how they built this social discovery tool, focusing on engineering challenges like prefetching for performance and using real-time signals for ML models. The conversation highlights how reducing the number of bubbles improved user engagement.
This episode explains how Meta built a feature showing friends' faces on Facebook Reels. Engineers discuss making it fast and how showing fewer friend circles actually made people use it more.
This episode offers a deep dive into the engineering behind a seemingly simple social feature. Listeners learn about scaling challenges, ML model evolution from survey data to real-time signals, and the iterative process of product development. It's for engineers and product managers interested in consumer social platforms and the practical application of ML in user experience. It connects to Meta's ongoing efforts to enhance social discovery and engagement within its core products.
- HostPascal Harris
- GuestJosephMeta
- GuestSubasreeMeta
- Episode Title: 85: Reel Friends: Building Social Discovery that Scales to Billions
- Show: Meta Tech Podcast
- Length: 39 minutes
- Feature Discussed: Friend Bubbles on Facebook Reels
- ML Model Change: Moved from survey-based friend rankings to real-time interaction signals
- Key Discovery: Showing fewer bubbles improved feature engagement
You've probably spotted those little circles of your friends' faces popping up on Facebook Reels. They look simple enough, but building them was a proper engineering challenge. In this episode, Pascal chats to Joseph and Subasree about Friend Bubbles, a feature that surfaces which of your close friends have been watching and reacting to the same Reels as you.
We get into the details of how prefetching keeps things snappy without wrecking scroll performance, why the team's ML model had to move from survey-based friend rankings to real-time interaction signals, and the surprising discovery that showing fewer bubbles actually made the whole feature click. If you've ever underestimated a "simple" feature, this one's for you.
Got feedback? Send it to us on Threads (https://threads.net/@metatechpod), Instagram (https://instagram.com/metatechpod) and don't forget to follow our host Pascal (https://mastodon.social/@passy, https://threads.net/@passy_). Fancy working with us? Check out https://www.metacareers.com/.
Links
Timestamps
-
Intro 0:06
-
Meet the Engineers: Backgrounds and Roles 1:53
-
Goals and Aspirations in Video Recommendations 4:20
-
The Origin of Friend Bubbles 4:41
-
Defining Success: Metrics and User Experience 5:40
-
Client-Side Constraints and Challenges 6:57
-
Feature Description: What Are Friend Bubbles? 8:31
-
Initial Challenges and Performance Issues 9:29
-
Architectural Changes for Performance 11:34
-
Impact of Performance on User Experience 15:14
-
Addressing Client-Side Challenges 16:58
-
Model Development: From Surveys to Interactions 20:07
-
Evolving the Model: Real-Time Data and User Interactions 23:35
-
Exploring Model Training and Performance 24:58
-
Feedback Loops and User Engagement 25:56
-
The Role of AI in Development 29:49
-
Collaboration Across Teams 32:17
-
Future Directions for Friend Bubbles 34:02
-
Safe Rollout Strategies for Features 35:22
-
Outro 37:31
-
Bloopers 38:27
- Prefetching
- Joseph and Subasree
- Loading content in advance to make it appear faster when you need it.
- ML model
- Joseph and Subasree
- A computer system trained to recognize patterns and make predictions or decisions.
38 min41 min87: Everything is Power: Battery Life Engineering for AI Glasses
WithRey, Dmitrii
43 min86: A Hard Cell: Engineering Ultra-Narrow Batteries for AI Glasses
WithKarthik, Myuran
1 h 22
43 min
22 min
Questions people ask
- What are Friend Bubbles on Facebook Reels?
- Friend Bubbles is a feature that shows which of your close friends have been watching and reacting to the same Reels as you, enhancing social discovery.
- What engineering challenges did the team face?
- The team focused on prefetching to maintain scroll performance and adapted their ML model to use real-time interaction signals instead of survey data.
Page generated Oct 19, 2026. Summaries are Subvolts' own; the audio, artwork and notes belong to Meta.