2026-06-03 17:00
🔌 Meta introduces the Instantaneous PowerLoss Storm, a new testing method to ensure data centers can handle sudden power losses. This approach enhances disaster preparedness by integrating resilience into existing systems with defense-in-depth strategies. It addresses challenges from zero-notice disasters, ensuring minimal impact on overall fleet availability. Learn more about how Meta is strengthening its infrastructure! ⚡🌐 #DataCenters #DisasterPreparedness #PowerLoss #MetaEngineering...
Source: Engineering at Meta
2026-05-26 16:00
🚀 Introducing SilverTorch, a new approach to recommendation systems that integrates all retrieval components into a unified architecture. With up to 23.7x higher throughput and 20.9x more compute cost efficiency than traditional methods, SilverTorch enhances user experience by delivering better recommendations quickly. This innovative model operates under the "Index as Model" paradigm, allowing for more complex and effective retrieval processes. Discover the full details in our research paper...
Source: Engineering at Meta
2026-05-13 13:00
🚀 The new Friend Bubbles feature in Facebook Reels showcases the Reels your friends have watched and engaged with. 🎙️ In the latest episode of the Meta Tech Podcast, engineers Subasree and Joseph discuss the complex engineering behind this seemingly simple feature. They share insights on the machine learning model and user behavior differences between iOS and Android. 🔗 Discover how small features can require significant effort. Listen now on Spotify, Apple Podcasts, or wherever you get your...
Source: Engineering at Meta
2026-05-12 16:00
🚀 Meta has revamped its data ingestion system to improve reliability at scale. The transition from a legacy system to a new architecture involved a comprehensive migration of all data ingestion processes. Key strategies included establishing a clear migration lifecycle and robust controls for a smooth transition. This upgrade enhances efficiency for analytics and product development across the company. #DataIngestion #MetaEngineering #TechInnovation #DataMigration #BigData
Source: Engineering at Meta
2026-05-11 16:00
🔒 Exciting news from Meta! Labyrinth 1.1 is here, enhancing the reliability of end-to-end encrypted backups for Messenger. This update introduces a new sub-protocol that ensures messages stay safe even if you switch devices or lose access for a while. Now, messages can be backed up as they are sent, improving the overall experience. For more details, check out the updated white paper. 📄 #DataSecurity #Encryption #Messenger #Meta #Privacy
Source: Engineering at Meta
2026-04-21 16:00
🚀 Facebook has revamped its Groups Search to enhance how users discover and validate community content. The new hybrid retrieval architecture improves search engagement by addressing three key friction points: discovery, consumption, and validation. Users can now find relevant information more easily, without facing the challenges of traditional keyword searches. This innovation aims to connect people better through shared interests. #FacebookGroups #CommunityKnowledge #SearchInnovation #Meta...
Source: Engineering at Meta
2026-04-16 16:00
🚀 Meta's Capacity Efficiency Program is transforming how we address performance issues through a unified AI agent platform. These agents automate the identification and resolution of performance problems, significantly reducing power usage and saving engineers' time for innovation. 💡 With tools like FBDetect, Meta catches thousands of regressions weekly, ensuring efficient resource management. The program aims for a self-sustaining efficiency engine, balancing proactive optimizations with...
Source: Engineering at Meta
2026-04-16 14:59
Meta has shared valuable lessons from its post-quantum cryptography (PQC) migration journey. This initiative aims to guide organizations as they adapt to emerging PQC standards. They introduce PQC Migration Levels to simplify the transition process across various use cases. Meta's framework covers risk assessment, inventory, deployment, and necessary safeguards. As quantum threats loom, organizations are encouraged to prioritize PQC protections. Meta contributes to this effort by co-authoring...
Source: Engineering at Meta
2026-04-09 16:00
At Meta, we enhance real-time audio and video experiences through WebRTC across multiple platforms. We faced challenges with forking this open-source project, risking isolation from community updates. To address this, we developed a dual-stack architecture, allowing A/B testing across 50+ use cases while ensuring continuous upgrades. This innovative approach has improved performance, binary size, and security for our services like Messenger and Instagram. Learn more about how we modernized...
Source: Engineering at Meta
2026-04-08 18:25
🎙️ In the latest episode of the Meta Tech Podcast, Pascal Hartig speaks with Ishwari and Joe from Meta’s Configurations team about ensuring safe config rollouts at scale. They discuss canarying, progressive rollouts, and the importance of health checks to catch issues early. The use of AI is also highlighted for reducing alert noise and improving response times. Listen now on your favorite podcast platform! 🎧 #MetaTech #AI #ConfigurationSafety #Podcast #TechInsights
Source: Engineering at Meta
2026-04-06 16:00
Meta has advanced its AI coding assistants by addressing limitations in understanding large-scale data pipelines. A project involving over 4,100 files and multiple languages revealed that AI agents struggled to make effective edits. To enhance performance, Meta created a pre-compute engine with over 50 specialized AI agents. This system generated 59 context files, capturing essential knowledge previously held by engineers. As a result, AI agents now cover 100% of code modules. The updated...
Source: Engineering at Meta
2026-04-02 19:59
🚀 Meta's Ranking Engineer Agent series continues with a focus on KernelEvolve, an innovative system designed to optimize AI infrastructure. This tool accelerates the performance of Meta's Ads Ranking models, enabling faster development and significant throughput improvements across various hardware. KernelEvolve streamlines the kernel optimization process, reducing weeks of work to hours and enhancing efficiency across multiple platforms. Learn more about this advanced approach to AI model...
Source: Engineering at Meta
2026-03-31 16:00
Meta is advancing its Ads Recommender models by implementing the Adaptive Ranking Model, designed to enhance AI Recommendation Systems. This innovation aims to improve understanding of user interests while maintaining efficiency. The model addresses the "inference trilemma," balancing complexity, latency, and cost. It utilizes intelligent request routing to serve the most effective model based on user context. Key developments include: - Inference-efficient model scaling for sub-second...
Source: Engineering at Meta
2026-03-30 16:00
Meta is advancing its efforts to support the construction industry with AI, aiming to enhance the production of sustainable concrete mixes made in the U.S. 🇺🇸 At the upcoming ACI Spring Convention, Meta will unveil BOxCrete, a new AI model for concrete mix design, along with foundational data for award-winning mixes. This open-source model is now available on GitHub. 🛠️ Despite the U.S. producing most ready-mix concrete, nearly 25% of cement is imported. Meta's initiative seeks to boost...
Source: Engineering at Meta
2026-03-18 18:19
🚀 Discover how Facebook Reels is enhancing social connections with "Friend Bubbles"! This new feature highlights Reels that your friends have liked or reacted to, making it easier to find content aligned with your interests. The system uses machine learning to assess relationship strength and rank video relevance, fostering meaningful engagement. By tapping on a bubble, you can even start direct conversations with friends about shared interests. Friend bubbles blend social signals with video...
Source: Engineering at Meta
2026-03-17 20:07
🚀 Meta introduces the Ranking Engineer Agent (REA), an autonomous AI agent that streamlines the machine learning lifecycle for ads ranking models. REA autonomously generates hypotheses, launches training jobs, and iterates on results, significantly reducing manual intervention. In its first rollout, REA doubled model accuracy and increased engineering output fivefold. Stay tuned for more insights on REA's capabilities! 🤖💡 #Meta #MachineLearning #AI #Innovation #Advertising
Source: Engineering at Meta
2026-03-13 16:00
🚀 Meta's Product Security team tackles the challenges of mobile security by developing secure-by-default frameworks for Android. They aim to make security updates easier for developers, using generative AI to automate code migration at scale. This innovative approach helps address vulnerabilities efficiently across vast codebases. Tune in to the latest episode of the Meta Tech Podcast to hear insights from the team! 🎧 #MobileSecurity #AI #MetaTech #Podcast #Engineering
Source: Engineering at Meta
2026-03-09 16:00
🛡️ Messenger's Advanced Browsing Protection (ABP) enhances user security by analyzing links shared in chats. It warns users about potentially malicious websites while maintaining privacy. 🔒 With end-to-end encryption, ABP employs a watchlist of dangerous sites and complex cryptographic systems to safeguard user information. 📊 The implementation uses private information retrieval techniques to minimize data exposure during link checks. #Messenger #AdvancedBrowsingProtection #Cybersecurity...
Source: Engineering at Meta
2026-03-02 20:00
Meta leverages FFmpeg for efficient media processing at scale. This tool supports various audio and video codecs, enhancing user experiences across apps. 🎥✨ Meta executes FFmpeg operations billions of times daily, addressing unique challenges in media file handling. Collaborating with FFmpeg developers, they have integrated essential features like multi-lane transcoding and real-time quality metrics. This collaboration has allowed Meta to move away from their internal fork and utilize the...
Source: Engineering at Meta
2026-03-02 17:00
Meta is renewing its commitment to jemalloc, a high-performance memory allocator crucial for its software infrastructure. The focus will be on reducing maintenance needs, modernizing the codebase, and adapting to new hardware and workloads. Collaboration with the open-source community is welcomed to enhance jemalloc's development. Meta aims to address technical debt and establish a long-term roadmap for this important project. #Meta #jemalloc #OpenSource #Infrastructure #SoftwareDevelopment 🚀🖥️💡
Source: Engineering at Meta
2026-02-24 21:30
🚀 Exciting news from Meta! We are open-sourcing RCCLX, an enhanced version of RCCL designed for AMD platforms. This new tool integrates with Torchcomms to support researchers in advancing AI innovations. Key features include Direct Data Access (DDA) and Low Precision Collectives, which significantly boost performance on AMD systems. Stay tuned for future updates as we expand this library! #OpenSource #AIInnovation #AMD #RCCLX #MetaTech
Source: Engineering at Meta
2026-02-11 17:00
The article discusses the evolution of software testing in light of agentic development. As code is created and deployed more rapidly, traditional testing methods struggle to keep up. Just-in-Time Tests (JiTTests) offer a novel solution. These tests are automatically generated by large language models to catch bugs immediately after code changes, without the need for constant maintenance. Catching JiTTests focus on identifying regressions and provide actionable feedback, allowing engineers to...
Source: Engineering at Meta
2026-02-09 17:00
🚀 Discover how Meta is advancing AI technology with backend aggregation (BAG) in its gigawatt-scale AI clusters like Prometheus. BAG connects thousands of GPUs across multiple data centers, ensuring seamless integration and high-capacity networking. By linking Disaggregated Schedule Fabric and Non-Scheduled Fabric, Prometheus aims to deliver 1-gigawatt capacity for enhanced AI experiences. This infrastructure will support future demands and drive innovation across Meta’s global network. #Meta...
Source: Engineering at Meta
2026-02-04 22:00
🚀 Exciting advancements in cross-device passkey authentication for XR devices are here! This new approach eliminates the need for QR codes, enabling secure connections for devices without displays. It adheres to trust and proximity requirements while enhancing usability. Developed in line with FIDO standards, this method is now available on Meta Quest devices, paving the way for secure, passwordless authentication across various screenless devices. 🔐 #CrossDeviceAuthentication #XRDevices...
Source: Engineering at Meta
2026-01-27 15:00
WhatsApp has introduced a new security layer for users, developed using Rust, to enhance defenses against malware. This initiative is part of their ongoing commitment to secure messaging for over 3 billion users. 🔒 The use of Rust addresses potential threats from malicious media files while maintaining operational efficiency. WhatsApp's past experiences, particularly with the 2015 Stagefright vulnerability, emphasized the need for robust protections. 🛡️ This rollout marks a significant step...
Source: Engineering at Meta
2026-01-14 20:51
🚀 Facebook Reels has enhanced its video recommendation system by focusing on user feedback instead of just likes and watch time. The new User True Interest Survey (UTIS) model aims to deliver niche, high-quality content, improving engagement and user satisfaction. This approach addresses challenges in personalization and seeks to better align content with users' true interests. Learn more about this innovative model and its impact on user experience in our detailed paper. 📄 #FacebookReels...
Source: Engineering at Meta
2026-01-12 18:34
🚀 CSS challenges arise with large websites, and Meta has responded by open-sourcing StyleX. StyleX merges the benefits of CSS-in-JS with the efficiency of static CSS. It offers atomic styling, reduces bundle size, and provides a user-friendly API for developers. Used by companies like Figma and Snowflake, StyleX is now standard across Meta's platforms like Facebook and Instagram. Listen to the Meta Tech Podcast for insights from software engineer Melissa, one of StyleX’s maintainers. 🎙️...
Source: Engineering at Meta
2025-12-22 14:00
📊 The 2025 Typed Python Survey reveals key insights into Python's type system and its adoption among developers. Conducted by JetBrains, Meta, and the Python community, it gathered 1,241 responses, showing a 15% increase from last year. 🔍 Most respondents are experienced developers, with nearly half having over a decade of Python experience. Reddit was the main platform for survey distribution, highlighting the diverse channels for developer engagement. 💡 Adoption of type hints remains...
Source: Engineering at Meta
2025-12-19 17:35
Meta recently introduced DrP, a root cause analysis platform that automates incident investigations. With over 300 teams using DrP, it performs 50,000 analyses daily, reducing the mean time to resolve incidents by 20-80%. Key features include an expressive SDK for creating analyzers, a scalable backend for executing them, and seamless integration with existing workflows. Learn more about how DrP enhances system reliability! 🔍💻⚙️ #Meta #RootCauseAnalysis #IncidentManagement #TechInnovation...
Source: Engineering at Meta
2025-12-17 14:00
🚀 Delve into the development of the Meta Ray-Ban Display, Meta's latest AI glasses! In a recent episode of the Meta Tech Podcast, Kenan and Emanuel from the Wearables team discuss the challenges faced in designing these innovative glasses. They explore topics from display technology to unique user interface patterns. Discover how particle physics relates to hardware design and the importance of celebrating small victories in a fast-paced environment. Listen now on your favorite podcast...
Source: Engineering at Meta
2025-12-15 17:00
Meta is advancing mobile security with its secure-by-default frameworks. These frameworks wrap potentially unsafe OS and third-party functions, ensuring security is a priority while keeping development fast and user-friendly. AI and automation are key in scaling these frameworks across Meta’s codebase, helping to identify insecure patterns and enforce consistent security measures. Balancing security, usability, and performance is crucial in framework design to encourage developer adoption and...
Source: Engineering at Meta
2025-11-21 21:00
🚀 Introducing Zoomer, Meta's new automated debugging and optimization platform for AI! Zoomer enhances performance across Meta's training and inference workloads, providing insights that lead to energy savings and improved workflow efficiency. It has significantly reduced training times and improved query per second (QPS) metrics. Operating at scale, Zoomer helps eliminate GPU underutilization, ensuring resources are used effectively, which in turn supports faster model iterations and product...
Source: Engineering at Meta
2025-11-20 17:00
🔒 Messenger has launched key transparency verification for end-to-end encrypted chats. This feature enhances security by ensuring only you and your contacts can access your messages. Key transparency provides an auditable record of public keys, helping users confirm their conversations are encrypted correctly. This reduces the need for manual key comparisons across devices. This advancement builds on industry efforts, similar to WhatsApp's implementation. #Messenger #KeyTransparency...
Source: Engineering at Meta
2025-11-18 17:00
🚀 Exciting news from Meta! We've launched Ax 1.0, an open-source platform designed for adaptive experimentation using machine learning. Ax helps improve AI models, optimize production infrastructure, and advance ML and hardware design. 🌟 Our new paper, “Ax: A Platform for Adaptive Experimentation,” details Ax’s architecture and compares its performance with other optimization libraries. Discover how Ax can enhance your research and experimentation process! #MachineLearning #OpenSource #AI...
Source: Engineering at Meta
2025-11-18 06:00
🚀 Exciting news! The core 2Africa infrastructure is now complete, marking the world’s longest open access subsea cable system. This landmark project connects East and West Africa, linking 33 countries across three continents, providing connectivity to over 3 billion people. 🌍 The initiative, led by Meta and global partners, aims to enhance digital experiences and foster economic growth in Africa and beyond. #2Africa #Connectivity #DigitalTransformation #SubseaCable #AfricaTech
Source: Engineering at Meta
2025-11-17 17:30
📱 Instagram has enhanced its iOS app by integrating Dolby Vision and ambient viewing environment (amve) support, improving the video experience for users. This update allows HDR videos from iPhones to maintain unique metadata, ensuring better color, brightness, and contrast across various displays. Instagram is now the first Meta app to fully support Dolby Vision, with plans for broader adoption across other Meta applications in the future. #InstagramUpdate #DolbyVision #VideoExperience #Meta...
Source: Engineering at Meta
2025-11-14 20:54
🌍 Open-source isn't just for software; it extends to hardware too! In the latest episode of the Meta Tech Podcast, Pascal Hartig discusses the environmental benefits of open hardware with Dharmesh and Lisa. They highlight Meta's initiatives from the 2025 Open Compute Project Summit, including a new AI methodology to tackle Scope 3 emissions. Discover how AI is aiding in the development of sustainable concrete for data center construction as Meta aims for net zero emissions by 2030. 🎧 Listen...
Source: Engineering at Meta
2025-11-11 17:58
🚀 StyleX is Meta's innovative styling system designed for large-scale applications. It merges the benefits of CSS-in-JS with the efficiency of static CSS, producing collision-free atomic CSS. Since its open-source launch in late 2023, StyleX has become the standard for Meta products like Facebook and Instagram. This compiler extracts styles at build time and facilitates intuitive styling across teams, enhancing predictability and scalability. #Meta #StyleX #CSStips #WebDevelopment #OpenSource
Source: Engineering at Meta
2025-11-10 17:00
🚀 Meta introduces the Generative Ads Recommendation Model (GEM), enhancing ad performance and advertiser ROI by improving ad relevance. GEM's architecture scales efficiently with more parameters, utilizing advanced GPU training techniques to deliver precise predictions. Since its launch, GEM has increased ad conversions by 5% on Instagram and 3% on Facebook Feed. This innovation marks a significant leap in ad recommendation systems. #Meta #GEM #AdsInnovation #AI #Advertising
Source: Engineering at Meta
2025-11-04 18:00
Meta is utilizing invisible watermarking to enhance content provenance on its platforms. This technology helps detect AI-generated videos, verify original posters, and track content sources. 🌐 The article discusses the challenges faced in scaling this watermarking solution, including the development of a CPU-based system that matches GPU performance while improving efficiency. 🔧 Invisible watermarking embeds data within media files, ensuring persistent identification even after edits or...
Source: Engineering at Meta