Reading room.
Engineering blogs & AI research
A reading room from engineering teams and AI labs: titles, short feed excerpts, and links to the originals. For my own essays, explore the writing archive.
- 1359 articles tracked
- 37 sources
- Refresh scheduled daily
Sources
Echoverse: Deep, evolving environments for computer-use agents
Computer-use AI agents struggle with multi-step workflows like email and customer support. Echoverse trains agents in realistic environments rather than simply providing more training tasks, helping them improve as the t…
EvoLib: Turning experience into evolving knowledge
LLMs do not get smarter just by remembering more. EvoLib turns experience into evolving knowledge, taking reusable skills and insights that help models learn and adapt across tasks long after deployment. The post EvoLib:…
GenRec: Towards LLM-Native Recommendation at Netflix
Authors: Ying Li , Arjun Rao , Shradha Sehgal Introduction Recommendations sit at the heart of the Netflix experience. Our current production models rely on thousands of hand‑crafted features over users, items, and inter…
How Salesforce Built an Agentic Engineering Enablement Strategy for Thousands of Software Engineers
Giving thousands of software engineers access to agentic tooling is straightforward. Helping them fundamentally change how they build software is another matter entirely; one every engineering organization eventually run…
From CUDA to MLX: How K-Search Brings Decades of Kernel Expertise to Apple Silicon
Figure 1: CUDA-to-MLX optimization translation map. CUDA optimization knowledge can be translated into architecture-native MLX strategies rather than copied instruction-for-instruction. We face a new epoch in computing.…
We’re launching Lyria 3.5 in Google Flow Music, with advances across musicality, lyrics, vocals, and creative control
Introducing Profile Frames: Decorative Borders to Make Your Discord Profile Museum-Worthy
Introducing Profile Frames: a new way to put a finishing touch on your Discord profile. Frames add a decorative border around your profile, giving the whole thing a look that s unmistakably yours.
Eval-driven development: Lessons from evaluating GenAI at scale
How Airbnb teams build trustworthy Generative AI products by treating evaluation as a first-class engineering discipline; not an afterthought. Nestled into the lush hillside, this stunning modern retreat features strikin…
What Are AI Agents? How They Work and Examples
Gemini Robotics 2 brings whole body intelligence to robots
AWS Weekly Roundup: Local Zone in Athens, Claude Opus 5 on AWS, Lambda durable execution for .NET, and more (July 27, 2026)
Last week I had the privilege of spending three days in São Paulo with technical builders from across Latin America, brought together for a regional tech event full of deep-dive sessions, hands-on workshops, and conversa…
Pinner Progression: Better Use-Case Representation Driving Weekly Active User Growth at Pinterest
Part 1 of 2 Authors Personalization (Homefeed): Yuke Yan, Chuxi Wang, Andreanne Lemay, Olafur Gudmundsson, Anna Kiyantseva, Krystal Benitez, Jongho Kim, Jiacong He, Rahul Goutam, James Li, Dylan Wang User Understanding:…
Building Reliable Production AI with Durable Workflows
Building an AI prototype has never been easier. You send a prompt to a model, receive a response, and present the result to a user. As long as each request is independent, the interaction is straightforward to reason abo…
Indexing the Data Lake for Online Point Queries
Companies like Spotify need vast quantities of data accessible at low latency for online services and,... The post Indexing the Data Lake for Online Point Queries appeared first on Spotify Engineering .
Teaching LLMs to Update Beliefs for Efficient Long-Horizon Interaction
.abbel-fig { display: block; text-align: center; margin: 2.4em 0; line-height: 1.4; max-width: 100%; } .abbel-fig img { display: block; margin: 0.65em auto 0; height: auto; max-width: 100%; } /* Image sizes; captions use…
Agent platform (Part 1): How we help Grab build and run AI agents at scale
Part 1: From one support bot to a framework At Grab, AI agents have evolved from interesting team prototypes into production services used every day by millions of merchants, drivers, and consumers. Today, more than 500…
The Microsoft 365 Copilot Agent’s Playbook: A Practical Livestream Series for Building Better Agents
Building on Microsoft 365 Copilot? Here s your playbook. Declarative agents are quickly becoming one of the most exciting ways to extend Microsoft 365 Copilot and bring organizational knowledge, workflows, and tools dire…
How AI Rebuilt Salesforce’s Decades-Old Localization Pipeline
In our Engineering Energizers Q A series, we highlight the engineering minds driving innovation across Salesforce. Today, we spotlight Teresa Marshall, Vice President of Localization. Teresa s team delivers every Salesfo…
Accelerating the frontiers of scientific discovery: Google’s $40M commitment to the Genesis Mission
Google commits $40M in AI tokens and credits for the Genesis Mission
Securing Infrastructure at Scale: Introducing Pinterest’s Resource Provisioner Pipeline (RPP)
Ammar Ekbote | Senior Software Engineer Chan Kim | Senior Software Engineer Managing Infrastructure as Code (IaC) across a massive organization comes with a unique set of security and logistical challenges, particularly…
Personalizing Airbnb search by learning from the guest journey
How we built a Transformer-based sequence model that encodes years of guest behavior to surface the right listings at the right time. By: Daochen Zha , Chun How Tan , Xin Liu , Bin Xu , Han Zhao , Xiaowei Liu , Jun Shi ,…
Introducing Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber
We’re introducing new Gemini models, including Gemini 3.6 Flash, 3.5 Flash-Lite and 3.5 Flash Cyber.
Introducing Gemini 3.6 Flash, 3.5 Flash-Lite, and 3.5 Flash Cyber
We’re introducing new Gemini models, including Gemini 3.6 Flash, 3.5 Flash-Lite and 3.5 Flash Cyber.
How to test agent experience changes without shipping them
Most changes you think will improve AI agent behavior won't. We tested a dozen hypotheses on a real project upgrade scenario and the majority failed. Learn how to emulate documentation, API, and MCP server changes locall…
How AI Reduced Customer Bug Triage from Nearly a Year to Less Than a Week
By Priya Sethuraman, Abhishek Ghose, Lovish Agarwal, and Aditya Pandey. In our Engineering Energizers Q A series, we highlight the engineering minds driving innovation across Salesforce. Today, we spotlight Priya Sethura…
Analyzing the evidence that helps businesses win “product not received” disputes
To understand what can influence win rates, we analyzed evidence packets from one million disputes over a 16-week period. Here’s what the data shows and what it means for how you mitigate disputes.
AWS Weekly Roundup: One-click Lambda setup prompt, OpenAI GPT-5.6 models on Bedrock, and more (July 20, 2026)
Last week, my team visited Seoul to meet AWS Korea User Group (AWSKRUG) leaders. AWSKRUG is the largest cloud developer community in Korea, with 20 meetup groups organized by topic and area that collectively host over 10…
How our universal content processing platform Riviera evolved for AI and beyond
Riviera is the Dropbox content processing platform that’s been iteratively improving content transformation in our products for roughly a decade.
Content Ingestion & Podcast Video Incident Report
Over the past two months, podcast creators have experienced a series of reliability issues on Spotify. This... The post Content Ingestion Podcast Video Incident Report appeared first on Spotify Engineering .
Introducing Gemini 3.5 Flash Cyber
Google introduces Gemini 3.5 Flash Cyber, a lightweight cybersecurity model to find and patch vulnerabilities.
How to test agent skills without hitting real APIs
Your agent skill calls an API. The moment you start evaluating it, every run either costs money or mutates production data. Learn how to mock APIs transparently so you can run evals without changing your skill or hitting…
In-House LLM Serving at Netflix
By AI Platform’s Model Runtime team and Inference team Introduction Most organizations consume LLMs through hosted APIs. Netflix went further — we run the full stack ourselves, from model deployment through inference, in…
Closing the Loop: How to Build Self-Improving AI Systems with Automated Feedback Loops
After six cycles, the system ran out of things to fix. That sentence probably raises more questions than it answers. Here is how we got there. The first skill PR took three days to merge. Twelve comments on structure. Si…
Our approach to bioresilience
Google DeepMind and Isomorphic Labs are sharing our joint approach to bioresilience and AI models.
Using Claude to Build an AI Knowledge Base in 30 Minutes
Ever had an experience onboarding someone, where you had to provide them a handful of documents, all authored at different times, and in various states of completeness? Ever had your AI agent burn cycles pulling down tha…
Migrating from Apollo Tooling to GraphQL Codegen at Yelp
Introduction At Yelp, we rely heavily on GraphQL and Apollo for data loading in our frontend React monorepo. When a developer writes a GraphQL query or mutation inside a React component, the shape of the response is defi…
Exploring Hierarchical Interest Representation For Meta Ads Deep Funnel Optimization
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 embeddi…
Building AX evals that actually work
This is the eighth and final article in a series about Agent Experience (AX): the practice of making AI coding agents work correctly with your technology. The series covers what you can and can t control in the agent sta…
Building AX evals that actually work
This is the eighth and final article in a series about Agent Experience (AX): the practice of making AI coding agents work correctly with your technology. The series covers what you can and can t control in the agent sta…
Robostral Navigate: single-camera AI navigation
From weeks to a day: how we made LLM evaluation fast enough to iterate on
Training an LLM is the easy part. The hard part is designing experiments and evaluations that you can trust enough to know whether the new model is actually an improvement. By : Baharak Saberidokht Introduction Shipping…
Blueberry: Force Multiplier For The On-Call Engineer
How we built a Slack-native on-call reasoning harness at Instacart that shortens time to first insight, speeds up theory testing, and turns tribal knowledge into reusable infrastructure. Key Contributors: Karthik Halukur…
Shipyard: How We Built Slack’s Next-Generation EC2 Platform
Over the past few years, we’ve been on a journey to modernise how we run Amazon Elastic Compute Cloud (EC2) instances at Slack. In our first post, Advancing Our Chef Infrastructure, we shared how we moved from a single C…
Training Orchestrator: Unifying Model Training at Yelp
At Yelp, we train many machine learning models on different schedules. Applied machine learning teams all have their own set of Spark-based training batches, scripts, and configurations. Over time, these diverged, leadin…
Amazon SQS turns 20: Two decades of reliable messaging at scale
On July 13, 2006, we launched Amazon Simple Queue Service (Amazon SQS) as one of the first three services available to customers, alongside Amazon EC2 and Amazon S3. We had learned firsthand that distributed systems need…
AWS Weekly Roundup: AWS Builder Center at 1 year, Network Scanning in Security Hub, Loom for AWS, and more (July 13, 2026)
AWS Builder Center turned one year old last week. Launched on July 9, 2025, the platform has grown from a community hub with Wishlist voting, community profiles, and a toolbox into a full ecosystem with sandbox environme…
Empowering India’s next generation of innovators with ATL Saathi
Google and AIM launched ATL Saathi, a Gemini-powered AI tool empowering Indian educators in robotics labs.
Modernizing the Meta Ads Service With an Open-Source Kernel Scheduler
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…
Verifying Rust cryptography in SymCrypt, from standards to code
Cryptographic code supports vital protections in modern computing systems. Learn how a new method helps verify code as developers write it while preserving speed and adaptability as it gets implemented and evolves. The p…
Building Service Topology at Scale: Architecture, Challenges, and Lessons Learned
By Parth Jain , Rakesh Sukumar , Yingwu Zhao , Renzo Sanchez-Silva Nathan Fisher A deep dive into the engineering challenges of building a real-time service dependency map at Netflix scale: from streaming architectures a…