---
title: "Latest reads"
description: "The engineering internet, summarised so you can actually read it."
---

# Latest reads
> The engineering internet, summarised so you can actually read it.

## Articles

### [Jun Kim, oMLX creator and maintainer, joins Hugging Face to support the MLX community](https://yomu.fyi/post/jun-kim-omlx-creator-and-maintainer-joins-hugging-face-to-support-the.md)
- Company: [Hugging Face](https://yomu.fyi/company/hugging-face.md)
- Author: Pedro Cuenca, Lysandre, Victor Mustar, Julien Chaumond, Jun Kim
- Published: Sep 22, 2026
- Type: Announcement

Jun Kim, creator and maintainer of oMLX, has joined Hugging Face to support the MLX community. MLX is Apple’s framework for local AI, optimized for Apple Silicon, and Hugging Face is the Hub for MLX models. oMLX, staying Apache 2.0, will gain stability and faster development as a funded, maintained project, allowing Jun to guide contributors and build for the long term. The goal is to unblock the community to run local AI in any form, with oMLX serving as a testbed leveraging dependencies like mlx-lm and mlx-vlm, and collaborating with projects such as LMStudio. A key focus is streamlining the transition from transformers model definitions to reference MLX implementations for different engines, so each can focus on unique features.


### [Simplifying Model Serving Across Multiple GPUs with NVIDIA TensorRT Multi-Device Integration in NVIDIA Dynamo-Triton](https://yomu.fyi/post/simplifying-model-serving-across-multiple-gpus-with-nvidia-tensorrt-mu.md)
- Company: [NVIDIA](https://yomu.fyi/company/nvidia-developer-blog.md)
- Author: Tanya Lenz
- Published: Sep 21, 2026
- Type: Announcement

NVIDIA TensorRT multi-device inference enables a single TensorRT network to execute across multiple GPUs using NCCL-backed distributed collectives, supported from TensorRT 11.0. Integrated into Dynamo-Triton release 26.07, one model instance can own multiple GPUs and expose a single gRPC endpoint so the client does not coordinate GPU ranks. The post demonstrates this with Cosmos 3 Nano video generation, where the transformer uses Ulysses context parallelism to distribute 44,160 video tokens across up to eight GPUs. End-to-end latency drops from 156.6 seconds on one GPU to 34.2 seconds on eight GPUs, achieving a 6.09x transformer RPC speedup. Teams can trade additional GPU resources for shorter request latency while keeping the application interface and surrounding workflow stable.


### [How to Evaluate AI Agents From Tool Calls to Task Completion](https://yomu.fyi/post/how-to-evaluate-ai-agents-from-tool-calls-to-task-completion.md)
- Company: [NVIDIA](https://yomu.fyi/company/nvidia-developer-blog.md)
- Author: Elizabeth Goodman
- Published: Sep 21, 2026
- Type: Explainer

Standard LLM benchmarks are insufficient for evaluating AI agents because they score static outputs rather than multi-step tool-call sequences against a live environment. The post explains that agent evaluation requires two scoring layers: step-level (process scoring) checks each call’s validity and usefulness, while end-to-end (E2E) scoring checks only the final environment state. A trace logs every step and environment state, and metrics roll up through a hierarchy of benchmark, trial, task, turn, and step. Key metrics include task success rate, consistency, tool-call precision, argument accuracy, steps per success, and cost per success. The post argues that benchmarks are converging on tool use because deployed agents always use tools, and enterprise benchmarks should measure task completion against real APIs and policies.


### [Amazon launches research initiative with Stanford University to advance AI and science](https://yomu.fyi/post/amazon-launches-research-initiative-with-stanford-university-to-advanc.md)
- Company: [AWS](https://yomu.fyi/company/aws.md)
- Author: Staff writer
- Published: Sep 21, 2026
- Type: Announcement

Amazon and Stanford University have launched the Stanford and Amazon Research Initiative, a formalized framework to advance research in AI, energy, and healthcare. The initiative builds on a deep relationship where currently more than 10 teams across Amazon fund active research and PhD fellowships at Stanford, spanning humanoid robotics, post-quantum cryptography, causal measurement science, and AI-driven radiology. It aims to accelerate the path from breakthrough research to real-world solutions that benefit society at large and broaden participation from diverse scholars. Focus areas will leverage strengths in artificial intelligence, machine learning, automated reasoning, and health, supported by Amazon’s cloud computing and AI services. The collaboration includes joint research projects, PhD fellowships, and interdisciplinary symposia, and was announced at a Stanford event with AWS CEO Matt Garman and Vice Provost David Studdert discussing the impact of university-industry collaborations.


### [Advancing AI for biology: Teaching models to design and characterize antibodies](https://yomu.fyi/post/advancing-ai-for-biology-teaching-models-to-design-and-characterize-an.md)
- Company: [AWS](https://yomu.fyi/company/aws.md)
- Author: Melih Yilmaz, Luca Giancardo, Yue Zhao, Edward Lee, Chuanyui Teh, Fangda Xu, Gordon Trang
- Published: Sep 21, 2026
- Type: Announcement

Monoclonal antibodies are a mainstay of biopharmaceutical development, with over 100 FDA-approved drugs and established manufacturing and regulatory pathways, yet conventional discovery is costly and slow, typically taking six to twelve months from target to lead candidate. AI promises to make antibody development cheaper, faster, and more flexible by designing and characterizing therapeutic antibodies computationally, but scientific questions remain. Development hinges on three factors: the best binding site on the target, which candidates bind most tightly, and whether candidates can survive manufacturing and the clinic. Predictive models for each factor do well on familiar targets and assays but considerably worse on unfamiliar ones, a gap that in-distribution benchmarks have made difficult to measure. Three papers from Amazon Bio Discovery’s science team address research questions about each factor, with methods integrated into the AI-powered application that connects biological AI models with lab services.


### [Open-Sourcing Rebalancer: A Generic, High-Performance Library for Solving Assignment Problems](https://yomu.fyi/post/open-sourcing-rebalancer-a-generic-high-performance-library-for-solvin.md)
- Company: [Meta](https://yomu.fyi/company/meta.md)
- Author: By Richard Barnes, Neeraj Kumar, Pol Mauri Ruiz
- Published: Sep 21, 2026
- Type: Announcement

Meta has open-sourced Rebalancer, a generic library for solving assignment problems that has been used internally for over nine years. Rebalancer separates problem specification from solution, using a three-step approach with modeling constructs like dimensions, partitions, scopes, and utilization, and a high-level spec API for common objectives and constraints. Problems are translated into an expression graph, and Rebalancer offers two solving techniques: an optimal solver using mixed integer programming (MIP) with commercial or open-source solvers, and a local search solver that evaluates millions of moves per second for large-scale problems. The local search solver handles problems with over a million objects and 5k bins, with an average solve time of 171 seconds. Rebalancer is used for roughly 40 million assignment problems daily across Meta’s infrastructure, and the open-source release includes a debugging UI called Rebalancer Explorer. The accompanying paper was published at OSDI’24.


### [Postgres on NVMe: performance and the convergence of transactions and analytics](https://yomu.fyi/post/postgres-on-nvme-performance-and-the-convergence-of-transactions-and-a.md)
- Company: [Clickhouse](https://yomu.fyi/company/clickhouse.md)
- Published: Sep 21, 2026
- Type: Benchmark

The post examines how local NVMe storage changes Postgres performance, particularly when working sets exceed memory and disk latency becomes the bottleneck. It argues that while faster storage accelerates transactional workloads, it does not alter the row-store layout, so analytical queries still require different approaches. Benchmark results show 9.2× more throughput on NVMe, with a median of 16,030 TPS versus 1,734 on EBS. The post introduces WalShadow, an open-source replication engine that reads the physical Postgres WAL to enable sub-second replication to ClickHouse without logical slots. It concludes that failover is solvable with tools like Patroni, and that storage latency is only part of the performance problem.


### [Turn Your Latest Observations Into Timely Weather Decisions With NVIDIA Earth-2](https://yomu.fyi/post/turn-your-latest-observations-into-timely-weather-decisions-with-nvidi.md)
- Company: [NVIDIA](https://yomu.fyi/company/nvidia-developer-blog.md)
- Author: Elizabeth Goodman
- Published: Sep 21, 2026
- Type: Tutorial

Weather-sensitive industries can now turn real-time observations into timely decisions using the AI data assimilation tools in NVIDIA Earth-2. The tutorial covers two techniques: Score-Based Data Assimilation (SDA) for constraining diffusion models with point observations, and HealDA for assimilating disparate datasets into a consistent global state. SDA works within downscaling models like CorrDiff and forecasting models like StormCast, guiding denoising steps toward observations without retraining. A concrete example showed SDA reducing wind-speed RMSE by 54% at held-out stations when applied to CorrDiff downscaling over Europe. Earth2Studio provides a unified interface across gridded, conventional, and satellite data sources, making it easy to connect proprietary observations to weather pipelines.


### [Introducing Our New Dropbox API Documentation](https://yomu.fyi/post/introducing-our-new-dropbox-api-documentation.md)
- Company: [Dropbox](https://yomu.fyi/company/dropbox.md)
- Author: Dropbox Developer Support Team
- Published: Sep 21, 2026
- Type: Announcement

Dropbox has launched a modernized API documentation site at https://docs.dropboxapi.com, rebuilt to make integration faster and more intuitive. The new documentation features interactive API testing that lets developers set parameters, issue API calls, and view real responses from the Dropbox API directly in their browser without switching tools. An embedded AI assistant offers instant help with endpoint questions, troubleshooting, and code examples, while the documentation integrates with MCP (Model Context Protocol) so AI tools like Claude Code and Cursor can access up-to-date API information. Through the MCP server at https://docs.dropboxapi.com/\_mcp/server, developers can connect their AI assistants to answer questions, explore the API, and assist with integrations using information directly from the docs. The update delivers a clean, responsive, searchable interface that brings endpoints, types, schemas, and related documentation together, with improved navigation, and provides comprehensive information including detailed error responses, type definitions, and full request/response schemas.


### [Pruning LLMs Like a Physicist: Block Removal as an Ising Optimization Problem](https://yomu.fyi/post/pruning-llms-like-a-physicist-block-removal-as-an-ising-optimization-p.md)
- Company: [Hugging Face](https://yomu.fyi/company/hugging-face.md)
- Author: Antonio Tiene, Ali Hashemi, David Jansen, Roman Rausch
- Published: Sep 21, 2026
- Type: Problem & solution

Removing entire transformer blocks is a cheap way to speed up large language models, yielding predictable inference gains and stacking with quantization and other compression methods. The challenge is that block choices interact, making it a combinatorial problem rather than a simple ranking task. The authors reformulate block selection as a constrained binary optimization (CBO) problem, mapping it onto an Ising glass, a disordered spin system with all-to-all interactions and a fixed number of removed blocks. They note that mean-field methods, which treat each block independently, are inadequate for this interacting system. Block removal composes with quantization, low-rank compression, width pruning, and knowledge-distillation-based healing, fitting into a larger pipeline rather than competing with it. The paper is titled ‘LLM Compression by Block Removal with Constrained Binary Optimization.’


### [Comparison of Cluster Autoscaler and KEDA MachineSet Autoscaling on Red Hat OpenShift](https://yomu.fyi/post/comparison-of-cluster-autoscaler-and-keda-machineset-autoscaling-on-re.md)
- Company: [Red Hat](https://yomu.fyi/company/red-hat.md)
- Author: Ramon Gordillo Gutierrez, Jose Ortiz Padilla
- Published: Sep 21, 2026
- Type: Explainer

This article compares two autoscaling approaches for Red Hat OpenShift compute infrastructure: the built-in Cluster Autoscaler and the KEDA-driven MachineSet Autoscaler. The key difference is the trigger mechanism: Cluster Autoscaler reacts to Pending pods, while KEDA scales based on external metrics like PromQL queries. KEDA triggers faster, scaling within about 15 minutes versus 24 minutes for Cluster Autoscaler, because it acts on metric thresholds before pods become unschedulable. Cluster Autoscaler offers built-in scale-down safety and zero configuration, while KEDA provides fine-grained, metric-driven control but requires the Custom Metrics Autoscaler operator. The choice depends on workload predictability, with Cluster Autoscaler suited for general-purpose workloads and KEDA for dedicated pools with predictable metric patterns. The two autoscalers cannot coexist on the same MachineSet, as they would fight over replica counts.


### [Scaling localnet user-defined networks (CUDNs) in Red Hat OpenShift to 3,500 networks](https://yomu.fyi/post/scaling-localnet-user-defined-networks-cudns-in-red-hat-openshift-to-3.md)
- Company: [Red Hat](https://yomu.fyi/company/red-hat.md)
- Author: Joe Talerico
- Published: Sep 21, 2026
- Type: Benchmark

Migrating from legacy virtualization with flat VLAN-based tenant networks to Red Hat OpenShift Virtualization often requires re-architecting network topology, creating friction. Red Hat OpenShift supports localnet user-defined networks (CUDNs) to preserve existing VLAN infrastructure while modernizing. Red Hat tested localnet CUDNs at scale, successfully managing up to 3,500 localnet CUDNs hosting over 3,500 active virtual machines. OVN-Kubernetes achieved 100% SSH connectivity, near-line-speed data-path performance over 100GbE interfaces without Geneve tunnel overhead, and stable control plane resource usage. Resource peaks reached approximately 5.5 CPU cores per worker node and 10 CPU cores on control plane nodes during bulk creation, then returned to low baselines.


### [AI threat modeling shouldn't be a bottleneck, so we open sourced ours](https://yomu.fyi/post/ai-threat-modeling-shouldn-t-be-a-bottleneck-so-we-open-sourced-ours.md)
- Company: [Red Hat](https://yomu.fyi/company/red-hat.md)
- Author: Katie Mulliken, Jon Weiser, Judy Kelly
- Published: Sep 21, 2026
- Type: Announcement

The post announces the open-sourcing of agentic-threat-modeling, an AI-assisted threat-modeling module for Claude Code, Cursor, and Gemini CLI, built to address the bottleneck of security teams reviewing AI-generated code. It runs on Lola, an open source package manager for AI contexts, and profiles likely attackers, generates narrative attack scenarios grounded in actual code, and checks them against six structured threat modeling frameworks (STRIDE, PASTA, LINDDUN, attack trees, and more). The module ties every finding to a specific file and line of code, as shown in a real-world example against a file upload endpoint. It’s designed to produce structured output for AI-assisted triage, though that pipeline isn’t built yet. The team open sourced it because the problem isn’t unique to them, and it’s available in RedHatProductSecurity/agentic-threat-modeling and LobsterTrap/lola.


### [Python Workers are now generally available](https://yomu.fyi/post/python-workers-are-now-generally-available.md)
- Company: [Cloudflare](https://yomu.fyi/company/cloudflare.md)
- Author: Hood Chatham
- Published: Sep 21, 2026
- Type: Announcement

Cloudflare announced general availability of Python Workers, making Python a first-class language on its developer platform. The runtime now encapsulates type conversion so developers use Cloudflare bindings like D1 and Workers AI directly, without the previously required JavaScript glue code. Popular frameworks FastAPI, Django, and Flask run inside Workers via built-in ASGI and WSCI connectors that translate JavaScript requests into standard Python structures. A custom socket syscall implementation enables TCP connections for database drivers, unlocking Hyperdrive integration with PostgreSQL and MySQL, previously blocked by WebAssembly sandbox limits. Example projects include streaming from Bluesky Jetstream via Durable Objects, building an MCP server with the official Python MCP package, and creating a RAG system with Vectorize, with documentation now including Python examples alongside TypeScript.


### [Inside Petal: Building the World’s First Petabit-Class Transoceanic Subsea Cable](https://yomu.fyi/post/inside-petal-building-the-world-s-first-petabit-class-transoceanic-sub.md)
- Company: [Meta](https://yomu.fyi/company/meta.md)
- Author: By Elizabeth Rivera Hartling, Pascal Pecci, Matthew Mitchell
- Published: Sep 21, 2026
- Type: Announcement

Meta announced Petal, the first transoceanic subsea cable to deliver petabit capacity, connecting France and the United States over 7,000 km and expected to enter service in 2029. It will be the first subsea cable system to deploy multi-core fiber technology at scale, using 2-core fiber in a 24 fiber-pair system to achieve 1 Pbps. The cable overcomes the Shannon Limit by adopting spatial division multiplexing with multi-core fiber, requiring ultra-pure synthetic silica for low attenuation and counter-propagation to minimize crosstalk. Petal’s repeaters use a single-body 96 amp design with a Fan-In/Fan-Out interface to transition between 2-core and single-core fiber for amplification, staying within existing 18 kV power feeding limits. Built with NEC, Sumitomo Electric, and Orange, Petal represents the largest generational capacity increase of any repeatered subsea system.


### [Building Sentry's Laravel AI Integration](https://yomu.fyi/post/building-sentry-s-laravel-ai-integration.md)
- Company: [Sentry](https://yomu.fyi/company/sentry.md)
- Author: Martin Linzmayer
- Published: Sep 21, 2026
- Type: Announcement

Sentry’s Laravel AI integration, built during an internal Agent Hackweek, provides Agent Tracing for Laravel AI agents. The author used Claude for coding, which initially produced poor generic helper code. Laravel AI emits lifecycle events for agents and tool calls, but not for individual LLM invocations. To capture these, the integration matches outgoing HTTP requests to active invocations using provider URL prefixes, associating with the most recently started invocation if multiple match. Enabling the integration requires only updating to sentry-laravel 4.27, with no Sentry-specific code. Traces appear in Sentry’s Explore > Agents section, showing conversations, tool calls, and their inputs and outputs.


### [Supporting a DPDPA-Compliant Ecosystem in India with MongoDB](https://yomu.fyi/post/supporting-a-dpdpa-compliant-ecosystem-in-india-with-mongodb.md)
- Company: [MongoDB](https://yomu.fyi/company/mongodb.md)
- Published: Sep 21, 2026
- Type: Explainer

India’s Digital Personal Data Protection Act (DPDPA), made law in 2023 and enforced in 2025, requires organizations to implement technical controls for consent, access, security, retention, and deletion of personal data. Traditional relational architectures often fragment customer information across multiple tables and services, making it difficult to track consent or respond to correction and deletion requests. MongoDB’s capabilities, including its flexible document model, change streams, queryable encryption, and search, can help consolidate privacy-relevant data and context. The document model allows storing consent state, personal data, and security controls together, reducing fragmentation and simplifying auditability. MongoDB Atlas and Enterprise Advanced offer deployment options for fully managed or self-hosted environments to support privacy-by-design architectures.


### [Sending SMS with an Agentic AI using Twilio and Hermes Agent](https://yomu.fyi/post/sending-sms-with-an-agentic-ai-using-twilio-and-hermes-agent.md)
- Company: [Twilio](https://yomu.fyi/company/twilio.md)
- Author: Amanda Lange
- Published: Sep 21, 2026
- Type: Tutorial

The author, a Twilio employee, built a personal agentic AI using Hermes Agent and connected it to Twilio SMS to enable phone-based interaction. They chose SMS over a custom Android app because it works out of the box on any phone and globally, making the agent accessible to anyone with the phone number, including non-technical users. The author wired up a toll-free 800 number and tested the setup by sending a text message from a grocery store parking lot, receiving a reply from the agent. The experiment demonstrated that agentic AI can be integrated with legacy protocols like SMS to provide realtime chat capabilities. The author also noted the importance of reading email provider terms of service if creating a new account for the agent.


### [More data won't make your AI smarter (here's what will)](https://yomu.fyi/post/more-data-won-t-make-your-ai-smarter-here-s-what-will.md)
- Company: [Twilio](https://yomu.fyi/company/twilio.md)
- Author: Jesse Sumrak
- Published: Sep 21, 2026
- Type: Problem & solution

Twilio research found that 99% of companies report something blocks them from understanding customers, with 48% citing connecting data across channels as the top obstacle. The problem isn’t data collection, which has succeeded, but the missing data activation layer that makes context usable in the eight seconds an AI agent has before a customer gives up. Twilio’s solution involves three elements: continuity across channels, real-time profile freshness, and retrieval of relevant history mid-conversation. Twilio Conversations treats touchpoints as a single lifecycle, Conversation Memory syncs new details in real time, and Conversation Intelligence surfaces approved documents alongside history. Twilio recommends starting small by picking one handoff where customers repeat themselves and making context survive it, rather than feeding the AI more data.


### [Automating Twilio Auth Token Detection and Rotation with TruffleHog, Tines, and Slack](https://yomu.fyi/post/automating-twilio-auth-token-detection-and-rotation-with-trufflehog-ti.md)
- Company: [Twilio](https://yomu.fyi/company/twilio.md)
- Author: Matt Coser
- Published: Sep 21, 2026
- Type: Tutorial

Hardcoding Twilio Auth Tokens in source code or exposing them in plaintext leaves communication systems vulnerable to account takeover, toll fraud, and unauthorized exploitation. This tutorial presents an automated detection and rotation pipeline using TruffleHog for secret scanning, Tines for security workflow automation, and Slack for user interaction. TruffleHog scans local directories or remote Git repositories for Twilio secrets, performs live verification via API requests, and outputs results in JSON. Tines ingests detection events through a webhook, parses the results, and triggers a Slack notification with interactive buttons for acknowledgment, scheduling remediation, or immediate rotation. Upon rotation, Tines calls the Twilio Account API to generate a new secondary auth token, displays it in a Slack modal for secure copy, and eventually promotes it to primary, while the old token is revoked.


[Newer posts](https://yomu.fyi/posts/page/37.md) · [Older posts](https://yomu.fyi/posts/page/39.md)