# Engineering at Meta Engineering at Meta Blog This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles. ## ZGateway: Learnings from Putting a Proxy in Front of ZippyDB DevFeed: [ZGateway: Learnings from Putting a Proxy in Front of ZippyDB]() Original publisher: [Read original article]() Author: Rittik Banik; Yunhao Cao Published: 2026-09-03T16:00:20Z Content type: article Language: en Sources: [Engineering at Meta]() Topics: [networking]() Tags: [backend](), [core-infra](), [database](), [engineering](), [meta](), [operations](), [resilience](), [routing](), [tls]() ### AI overview Meta introduces ZGateway, a proxy layer that unifies client traffic to its ZippyDB key-value store. The article explains how a managed proxy centralizes connection management, request batching, load balancing, admission control, routing, and cross-region resilience for a large client fleet. ### Source excerpt We're introducing ZGateway, the proxy we are using to unify traffic through ZippyDB, Meta's most widely-used key value store. As a bonus, it also enables admission control, load balancing, cross-region resilience, and richer operations. ZippyDB is the most widely used key value store at Meta, backing product metadata, counters, and configuration, and can serve billions [...] Read More... The post ZGateway: Learnings from Putting a Proxy in Front of ZippyDB appeared first on Engineering at Meta. ## An Organizational Second Brain: Building an AI That Learns From Experts DevFeed: [An Organizational Second Brain: Building an AI That Learns From Experts]() Original publisher: [Read original article]() Author: Shaurya Sengar; Jason Nawrocki; Jay Shah; Prashant Kommireddi Published: 2026-09-02T09:00:29Z Content type: article Language: en Sources: [Engineering at Meta](), [Meta AI Research](), [Meta ML Applications]() Topics: [Large Language Model](), [Architecture & Design]() Tags: [ai-agent](), [ai-research](), [architecture](), [compliance](), [llms](), [ml-applications](), [security-privacy]() ### AI overview Meta describes an AI agent for a compliance domain that preserves specialist knowledge through an auditable knowledge architecture, an expert-like reasoning layer, and a feedback-driven improvement pipeline without model retraining. ### Source excerpt We've built an AI agent that acts as a secondary expert for a given domain, making deep specialist knowledge readily available and preserved for anyone in an organization to access, share, and build upon. This is not a typical domain-specific agent. Its novelty comes from integrating two layers: A structured, auditable knowledge architecture separates what [...] Read More... The post An Organizational Second Brain: Building an AI That Learns From Experts appeared first on Engineering at Meta. ## MetaRoCE: A New RDMA Transport Built for AI-Scale Ethernet DevFeed: [MetaRoCE: A New RDMA Transport Built for AI-Scale Ethernet]() Original publisher: [Read original article]() Author: Arvind Srinivasan; Neil Spring; Omar Baldonado; Rajiv Krishnamurthy Published: 2026-08-24T18:02:29Z Content type: article Language: en Sources: [Engineering at Meta]() Topics: [Ethernet](), [Networks](), [AI Infrastructure](), [GPU](), [data centers](), [Low Latency](), [telemetry](), [Inference]() Tags: [ai-infrastructure](), [data-center-engineering](), [data-centers](), [data-infrastructure](), [ethernet](), [gpu](), [low-latency](), [networking-traffic](), [open-source](), [performance](), [real-time](), [telemetry]() ### AI overview Meta introduces MetaRoCE, an RDMA transport protocol designed for AI workloads on commodity Ethernet at million-GPU scale. The article describes its release through the Open Compute Project and explains how endpoint intelligence, packet spraying, fine-grained logical paths, and real-time telemetry aim to provide high throughput, low tail latency, and operational simplicity for distributed training and inference. ### Source excerpt Training and serving frontier AI models depends on fast, reliable networks that move data between GPUs without wasting compute cycles. To meet this challenge at scale, Meta designed MetaRoCE - a clean-sheet RDMA transport protocol purpose-built for AI workloads on commodity Ethernet. We're releasing the MetaRoCE specification, a reference software implementation and a compliance test [...] Read More... The post MetaRoCE: A New RDMA Transport Built for AI-Scale Ethernet appeared first on Engineering at Meta. ## MTIA 300: Meta's First Training Chip with Built-in NICs and Communication-Offloading Engines DevFeed: [MTIA 300: Meta's First Training Chip with Built-in NICs and Communication-Offloading Engines]() Original publisher: [Read original article]() Author: Rajiv Krishnamurthy; Wes Bland Published: 2026-08-24T17:45:52Z Content type: article Language: en Sources: [Engineering at Meta]() Topics: [recommendation systems](), [Language models]() Tags: [chip-design](), [communication](), [data-infrastructure](), [devinfra](), [gpu](), [hardware](), [inference](), [infrastructure](), [meta](), [networking-traffic](), [performance](), [production-engineering](), [training]() ### AI overview Meta describes MTIA 300, an in-house accelerator for training ranking and recommendation models, with built-in network chiplets and a co-designed HCCL communication library. The design targets communication-heavy distributed training by integrating RDMA NICs into the chip package and offloading communication work. ### Source excerpt MTIA 300 is the first of Meta's family of in-house training and inference accelerators optimized for training ranking and recommendation models. We're sharing how MTIA 300's built-in NIC chiplets allow it to meet the communication needs associated with training recommendation models with superior performance over general-purpose GPUs. By co-designing MTIA's communication library, HCCL, alongside the [...] Read More... The post MTIA 300: Meta's First Training Chip with Built-in NICs and Communication-Offloading Engines appeared first on Engineering at Meta. ## How We're Building Scam Alert on WhatsApp With End-to-End Encryption and Verifiability Guarantees DevFeed: [How We're Building Scam Alert on WhatsApp With End-to-End Encryption and Verifiability Guarantees]() Original publisher: [Read original article]() Author: Chris Wiltz Published: 2026-08-12T13:00:28Z Content type: article Language: en Sources: [Engineering at Meta]() Topics: [End-to-End Encryption](), [Encryption](), [Machine Learning & Artificial Intelligence](), [Inference](), [Security](), [Bug Bounty](), [Social engineering]() Tags: [bug-bounty](), [encryption](), [machine-learning](), [privacy](), [scam](), [security](), [security-privacy](), [social-engineering](), [user-control](), [whatsapp]() ### AI overview An early technical overview of WhatsApp's optional Scam Alert feature, which uses a small on-device machine learning model to classify potential scam messages while keeping message content on the device. The article explains how the design preserves end-to-end encryption through local processing, avoids automatic reporting, gives users control, and supports independent security review during a limited Beta rollout. ### Source excerpt WhatsApp is committed to helping people stay safe while protecting the privacy of their messages. As scam tactics evolve -- from impersonation to social engineering to AI-generated lures -- we're always evolving as well, so that our protections stay ahead of scammers while protecting people's personal messages with end-to-end encryption. Today, we're sharing an early [...] Read More... The post How We're Building Scam Alert on WhatsApp With End-to-End Encryption and Verifiability Guarantees appeared first on Engineering at Meta. ## From User Sequences to Scaling Laws: A Multi-Stage Architecture for Meta's Ads Ranking DevFeed: [From User Sequences to Scaling Laws: A Multi-Stage Architecture for Meta's Ads Ranking]() Original publisher: [Read original article]() Author: Steven De Gryze; Parshva Doshi; Sean O'Byrne; Arnold Overwijk; Dinesh Ramasamy; Lee Xiong Published: 2026-08-05T19:20:20Z Content type: article Language: en Sources: [Engineering at Meta](), [Meta ML Applications]() Topics: [recommendation systems](), [Temporal data]() Tags: [ads](), [architecture](), [data-infrastructure](), [meta](), [ml-applications](), [production-engineering](), [recommendation-systems](), [scaling-laws](), [tokenization]() ### AI overview Meta describes a multi-stage sequence-model architecture for ads ranking that separates offline user modeling from lightweight online ranking. It also uses dense tokenization and target-aware attention to learn feature interactions, with reported conversion and ad-click lifts across Instagram and Facebook. ### Source excerpt Every day, Meta's recommendation platforms handle billions of user interactions, generating rich temporal signals that capture individual preferences and intent across products, ads, and content. In our 2024 post on sequence learning for ads recommendations, we showed how modeling the order and timing of user actions (rather than relying on static, manually engineered sparse features) [...] Read More... The post From User Sequences to Scaling Laws: A Multi-Stage Architecture for Meta's Ads Ranking appeared first on Engineering at Meta. ## GEM Training: How Meta Doubled the Efficiency of Its LLM-Scale Ads Foundation Model DevFeed: [GEM Training: How Meta Doubled the Efficiency of Its LLM-Scale Ads Foundation Model]() Original publisher: [Read original article]() Author: Darren Liu; Huayu Li; Raghav Boinepalli; Yuzhen Huang; Jackie (Jiaqi) Xu; Richard Qiu; Chunzhi Yang; Rich Zhu; Dev (Devashish) Shankar; Huaqing Xiong Published: 2026-08-03T18:00:17Z Content type: article Language: en Sources: [Engineering at Meta](), [Meta AI Research](), [Meta ML Applications]() Topics: [recommendation systems](), [networking]() Tags: [ads](), [ai-infrastructure](), [ai-research](), [data-infrastructure](), [gpu](), [hardware](), [kernels](), [llm](), [meta](), [ml-applications](), [networking](), [recommendation-systems](), [training]() ### AI overview Meta describes training its GEM ads recommendation foundation model at LLM scale. The article covers recommendation-specific kernels, ultra-low-precision training, and topology-aware parallelism that doubled end-to-end training efficiency to 20-25% MFU while increasing training FLOPs fourfold. ### Source excerpt Meta's Generative Ads Recommendation Model (GEM), the foundation model behind ads recommendations across Instagram and Facebook, now trains at LLM scale on several thousand of the latest-generation GPUs. This post goes into the details on how we achieved: doubling end-to-end (E2E) training efficiency to 20-25% Model FLOPs Utilization (MFU) while scaling training FLOPs 4x in [...] Read More... The post GEM Training: How Meta Doubled the Efficiency of Its LLM-Scale Ads Foundation Model appeared first on Engineering at Meta. ## Exploring Hierarchical Interest Representation For Meta Ads Deep Funnel Optimization DevFeed: [Exploring Hierarchical Interest Representation For Meta Ads Deep Funnel Optimization]() Original publisher: [Read original article]() Author: Yuhui Ouyang; Di Wang; Sreedal Menon; Jie Tian Published: 2026-07-15T17:00:52Z Content type: article Language: en Sources: [Engineering at Meta](), [Meta AI Research]() Topics: [Optimization](), [Embeddings](), [Graphs](), [recommendation systems](), [LLMs](), [Artificial Intelligence](), [Transformer](), [data]() Tags: [ads](), [ai](), [ai-research](), [data](), [embeddings](), [engineering](), [frontier-ai](), [generative](), [learning](), [meta](), [optimization](), [personalization](), [reasoning](), [research](), [retrieval](), [scale]() ### AI overview Meta describes Hierarchical Interest Representation, an upstream system that learns unified embeddings for users, advertisers, products, and services. It combines graph learning, multimodal content processed through LLMs, engagement signals, and self-supervised distillation to improve personalization, retrieval, ranking, and deep-funnel advertising optimization. ### Source excerpt Hierarchical Interest Representation is a research area for Meta Ads. We're exploring an upstream representation layer over the universe of Ads entities - users, advertisers, products, services - learning unified embeddings that connect users' inferred interests with the breadth of what advertisers offer in their deep funnel ads. The innovations in Hierarchical Interest Representation are [...] Read More... The post Exploring Hierarchical Interest Representation For Meta Ads Deep Funnel Optimization appeared first on Engineering at Meta. ## Modernizing the Meta Ads Service With an Open-Source Kernel Scheduler DevFeed: [Modernizing the Meta Ads Service With an Open-Source Kernel Scheduler]() Original publisher: [Read original article]() Author: Blaise Sanouillet; Sunyi Shao; Jeff Song; Gosh Arzumanyan; Shawn Wu; Tejun Heo; Praveen Sampath; GP Musumeci Published: 2026-07-13T16:00:50Z Content type: article Language: en Sources: [Engineering at Meta](), [Meta ML Applications]() Topics: [Latency](), [Linux](), [Optimization](), [Open Source](), [Google]() Tags: [engineering](), [kernel](), [latency](), [linux](), [meta](), [ml-applications](), [open-source](), [optimization](), [performance]() ### AI overview Meta describes using the open-source, BPF-based sched_ext framework to create a Linux kernel scheduling policy tailored to its ads delivery workload. The approach addressed latency regressions associated with a kernel upgrade and produced lower tail latency, reduced power consumption, and a modest increase in ads ranked. ### Source excerpt TL; DR At Meta's scale, a few milliseconds of latency degradation can have a significant negative impact on ads performance. When a Linux kernel upgrade risked regressing latency across Meta's ad serving fleet, we turned to sched_ext -- the upstream, BPF-based extensible scheduling framework -- to build a scheduling policy customized to the Ads delivery [...] Read More... The post Modernizing the Meta Ads Service With an Open-Source Kernel Scheduler appeared first on Engineering at Meta.