---
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

### [Introducing Governance Hub: Intelligent, account-level governance over your Databricks estate](https://yomu.fyi/post/introducing-governance-hub-intelligent-account-level-governance-over-y.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Sadhana Bala, Kristen Wilder, Jodi Boone
- Published: Aug 26, 2026
- Type: Announcement

Databricks introduces Governance Hub, a beta account-level interface for platform and governance teams overseeing estates spanning hundreds of workspaces and multiple regions. It consolidates data health, access, AI usage, and cost information across AWS, Azure, and GCP, replacing scattered system tables, workspace views, and third-party tools with drill-downs for metadata gaps, principal access, token and model activity, and untagged spend. The hub also integrates with Genie, allowing users to ask governance questions in natural language and receive answers and recommendations grounded in governance data; future actions such as configuring policies and alerts are described as planned. Account admins enable it from the Account Console, with access varying by role and existing permissions respected, while upcoming features include performance and security insights and delegated recommendations.


### [GRPO fine-tuning on Red Hat OpenShift AI: Reinforcement learning from verifiable rewards with Training Hub](https://yomu.fyi/post/grpo-fine-tuning-on-red-hat-openshift-ai-reinforcement-learning-from-v.md)
- Company: [Red Hat](https://yomu.fyi/company/red-hat.md)
- Author: Fiona Waters
- Published: Aug 26, 2026
- Type: Tutorial

Supervised fine-tuning for structured reasoning tasks like tool calling often demands expensive, labor-intensive dataset curation. Group Relative Policy Optimization (GRPO) resolves this by using programmatic reward functions to score multiple candidate responses without requiring human labels or a dedicated reward model. On Red Hat OpenShift AI, developers can implement GRPO using Training Hub, Kubeflow Trainer, and the Kubeflow SDK to fine-tune models like Qwen3-4B. The setup utilizes vLLM for candidate generation alongside LoRA weight updates, sharing storage via ReadWriteMany persistent volume claims. In a practical tool-calling test, five training iterations raised task accuracy from 33% to 67%, teaching the model to emit parseable function invocations rather than fabricated conversational text.


### [MongoDB Atlas Joins Google Cloud Foundation Fabric (FAST): Secure Private Data Infrastructure in Minutes](https://yomu.fyi/post/mongodb-atlas-joins-google-cloud-foundation-fabric-fast-secure-private.md)
- Company: [MongoDB](https://yomu.fyi/company/mongodb.md)
- Published: Aug 26, 2026
- Type: Announcement

Many organizations lack cloud fundamentals, making foundational setup complex and time-consuming; Google Cloud Foundation Fabric (FAST) is a GitOps-driven landing zone toolkit that accelerates establishing a production-ready Google Cloud organization. A new MongoDB Atlas module within FAST enables provisioning MongoDB Atlas clusters with Private Service Connect (PSC) for secure, private connectivity. PSC avoids VPC peering issues like CIDR conflicts and transitive routing, using a unidirectional consumer-producer model that respects project boundaries and routes traffic over Google’s private fiber network. The module automates cluster creation, PSC endpoint registration with Service Directory, and internal DNS resolution, integrating with Google Cloud KMS for customer-managed encryption keys. This integration reduces foundational setup from weeks to minutes, providing a secure data foundation that connects to BigQuery for real-time analytics and Vertex AI for GenAI workloads.


### [Enterprise-Grade Precision for Long-Context Multimodal Embedding Inference on Cloud TPU](https://yomu.fyi/post/enterprise-grade-precision-for-long-context-multimodal-embedding-infer.md)
- Company: [Google](https://yomu.fyi/company/google.md)
- Author: Anthony Su, Injae Kwak
- Published: Aug 26, 2026
- Type: Benchmark

Embedding models convert text, images, and audio into dense vectors used for semantic search, recommendation, intent classification, content discovery, and clustering. The post describes production support for long-context Qwen3 Embedding models on Cloud TPUs through native vLLM TPU integration and GKE Custom Compute Classes, allowing prioritized autoscaling across TPU and GPU capacity pools. Its implementation pads vocabulary tensors for hardware-safe sharding, hardens lazy loading, pre-warms sharding-aware JAX/XLA compilation caches, and uses a hybrid StepPool with CachedRequestState to preserve pooling state during chunked prefill and preemption. Parity evaluations target cosine similarity of at least 0.999 for text and 0.995 for multimodal inputs; on TPU Ironwood, Qwen3-Embedding-8B reached 83,996 total tokens per second and 5.13 requests per second at 16K+ sequence length with bf16 and TP=4.


### [Build a Video Chat Application with Python, JavaScript and Twilio Programmable Video](https://yomu.fyi/post/build-a-video-chat-application-with-python-javascript-and-twilio-progr.md)
- Company: [Twilio](https://yomu.fyi/company/twilio.md)
- Author: Miguel Grinberg
- Published: Aug 26, 2026
- Type: Tutorial

This tutorial builds a browser-based video chat application using Python, Flask, vanilla JavaScript, HTML, CSS, and Twilio Programmable Video. It adds text chat through Twilio Conversations and screen sharing, allowing participation on modern desktop and mobile browsers without installed software. The implementation covers Twilio account and API-key setup, environment variables, a Python virtual environment, client-side event handlers, and ngrok, which exposes the local Flask server through an HTTPS public URL for testing on other devices. The refreshed project targets Flask 3.x, twilio-python 9.x, Twilio Video JS 2.35.x, and Twilio Conversations classic JS 2.x, and the completed flow supports live calls, screen sharing, and text chat. Participants must use unique names because Twilio identifies them by token identity.


### [8 top conversational AI platforms in 2026](https://yomu.fyi/post/8-top-conversational-ai-platforms-in-2026.md)
- Company: [Twilio](https://yomu.fyi/company/twilio.md)
- Author: Jesse Sumrak
- Published: Aug 26, 2026
- Type: Explainer

Published in August 2026, this guide compares eight conversational AI platforms amid rapid market consolidation and acquisitions including NiCE’s roughly $955 million purchase of Cognigy. It defines the category as software for building, deploying, and operating agents across voice and messaging, then distinguishes vendor-managed services that own reasoning from infrastructure providers that carry conversations while teams own the reasoning. The evaluation criteria include model flexibility, voice and messaging parity, deployment control, cross-channel context, human handoff, pricing, and integration depth. Twilio is presented as an infrastructure layer for teams choosing their own models, with ConversationRelay, Agent Connect, Conversation Orchestrator, Conversation Memory, and Flex supporting channels, continuity, and escalation; other entries include Google Conversational Agents, Kore.ai, and NiCE Cognigy. The guide argues that portable models and conversation data will matter as the conversation layer separates from agent reasoning.


### [Leveraging Databricks to Support FISC Security Guidelines](https://yomu.fyi/post/leveraging-databricks-to-support-fisc-security-guidelines.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Jarrett Hoffman, Abhi Arikapudi
- Published: Aug 26, 2026
- Type: Explainer

The FISC Security Guidelines define control objectives for Japanese financial institutions, while the post addresses the challenge of translating them into auditable controls on Databricks data and AI workloads. The Databricks Data + AI Platform, Unity Catalog, and optional Enhanced Security Monitoring and Compliance Security Profile features are presented for identity and network controls, encryption, fine-grained access, audit logs, lineage, and hardened runtimes. For AI workloads, Unity Catalog’s Model Registry links training data, code versions, and evaluation metrics to model versions, while a separate framework maps 97 AI-specific risks to 73 controls. The FISC Customer Capabilities Mapping Matrix connects guideline units to platform capabilities and customer responsibilities, including configuration, SIEM integration, backup, recovery, and evidence; the Security Analysis Tool is suggested for finding workspace gaps, and the material is identified as educational guidance requiring expert review.


### [What's new in the ClickHouse .NET Driver: the road from 1.0 to 1.3](https://yomu.fyi/post/what-s-new-in-the-clickhouse-net-driver-the-road-from-1-0-to-1-3.md)
- Company: [Clickhouse](https://yomu.fyi/company/clickhouse.md)
- Published: Aug 26, 2026
- Type: Announcement

The ClickHouse .NET Driver moved from its 1.0 stable release to versions 1.1–1.3, changing how applications insert, read, and parameterize data while expanding type coverage, configurability, observability, and .NET ecosystem integrations. The largest ergonomic change is registered POCO support: RegisterPocoType() validates mappings up front, pre-compiles getters and setters, and enables InsertBinaryAsync with IEnumerable plus lazy QueryAsync() results through IAsyncEnumerable. Attributes customize column and ClickHouse type mappings, JSON serialization supports POCOs with paths and ignored properties when JsonWriteMode is Binary, and three optional extension points govern parameter type inference, wire formatting, and read conversion. The announced 1.4 roadmap targets allocation reductions, pluggable compression, and box-free POCO operations; version 1.5 is planned to add a Native/TCP client with columnar transfer and server progress information.


### [Training and Finetuning Multi-Vector Embedding Models with Sentence Transformers](https://yomu.fyi/post/training-and-finetuning-multi-vector-embedding-models-with-sentence-tr.md)
- Company: [Hugging Face](https://yomu.fyi/company/hugging-face.md)
- Author: Tom Aarsen
- Published: Aug 26, 2026
- Type: Tutorial

Sentence Transformers v6.0 adds MultiVectorEncoder, a fourth model type for ColBERT-style late-interaction retrieval, alongside a training approach for finetuning existing models or building them from base transformers. Multi-vector models retain one small vector per token and score query-document pairs with MaxSim, where each query token selects its best document-token match and the scores are summed. The guide covers models, datasets, loss functions, training arguments, evaluators, trainers, callbacks, and multi-dataset training, with examples using pip install -U "sentence-transformers\[train\]". On a medical retrieval evaluation, the finetuned multi-vector-encoder/mLateOn-medical model, trained for 14.5 hours on a single RTX 3090, outperformed every general-purpose retrieval model tested across dense, sparse, lexical, and multi-vector systems.


### [Building a data plane from scratch: Stripe’s own high-performance distributed proxy](https://yomu.fyi/post/building-a-data-plane-from-scratch-stripe-s-own-high-performance-distr.md)
- Company: [Stripe](https://yomu.fyi/company/stripe.md)
- Author: Jack Loughran
- Published: Aug 26, 2026
- Type: Problem & solution

Stripe replaced Envoy in its service mesh after reliability, connection-cardinality, rollout-risk, and customization problems became increasingly difficult at its scale. Mesh-proxy, a Go proxy, implements HTTP/1.1 and HTTP/2, connects directly to Stripe’s service registry, and supports retries, hedging, rate limiting, and dynamic priority-based routing. Unlike Envoy’s worker-per-core model, it uses one read and one write goroutine per connection and multiplexes requests over a shared connection, reducing connection counts from one per worker to one per host. Load tests and deployment results showed about half the CPU under high load and roughly 50% lower latency, equivalent to around 3,000 days of cumulative inter-service latency saved daily. Direct registry integration also reduced unnecessary xDS updates, while custom Ruby balancing added 15–20% headroom before bursts overwhelm machines; Stripe is now pursuing anomaly detection and smart routing.


### [Migrating to Auth0 for Startups from Another Auth Solution](https://yomu.fyi/post/migrating-to-auth0-for-startups-from-another-auth-solution.md)
- Company: [Auth0](https://yomu.fyi/company/auth0.md)
- Author: Shreya Gupta
- Published: Aug 26, 2026
- Type: Problem & solution

Evenpay develops compensation transparency software to meet the EU Pay Transparency Directive, requiring strict safeguards for sensitive employee salary data. The startup initially selected a lightweight B2B authentication platform for quick onboarding, but soon encountered unmet data residency and compliance demands from Finnish enterprise clients. To resolve these limitations, Evenpay migrated all customers to Auth0 over a few weeks using documentation and AI assistance. Adopting Auth0 allowed the team to fulfill regional EU data controls while reducing managed Single Sign-On setup times from weeks to as little as one day through self-service options. Furthermore, relying on an established authentication brand provided instant credibility with enterprise security officers without requiring custom authentication development.


### [PROOF-Gen: From Optimized Data to Better Distillation](https://yomu.fyi/post/proof-gen-from-optimized-data-to-better-distillation.md)
- Company: [Apple](https://yomu.fyi/company/apple.md)
- Published: Aug 26, 2026
- Type: Problem & solution

Standard distillation pipelines for tool-calling agents use generate-and-filter supervised fine-tuning on teacher trajectories, discarding failed trials and missing training signals on hard scenarios. On τ 2-bench, 57% of teacher trials fail, with two-thirds being near-misses undone by a single error. PROOF-Gen (Per-scenario Reflective Optimization to Overcome Failed Generation) addresses this loss by employing a reflector that reviews execution traces and evaluation feedback to craft corrective guidance for the teacher. Once passing trajectories are generated, this scaffold guidance is stripped so students train on clean demonstrations. Per-scenario prompt optimization recovers 93% of failed scenarios on τ 2-bench, boosting Qwen3-4B-Instruct-2507 Passˆ1 from 0.132 to 0.529 and lifting deployed on-device model performance across all locales.


### [IDEA Prune: An Integrated Enlarge-and-Prune Pipeline in Generative Language Model Pretraining](https://yomu.fyi/post/idea-prune-an-integrated-enlarge-and-prune-pipeline-in-generative-lang.md)
- Company: [Apple](https://yomu.fyi/company/apple.md)
- Published: Aug 26, 2026
- Type: Problem & solution

Recent advancements in large language models necessitate efficient deployment options within constrained inference budgets. While structured pruning offers better token efficiency than training target-sized models from scratch, previous work frequently ignores the pretraining of enlarged models. To resolve this gap, researchers developed an integrated enlarge-and-prune pipeline that unifies enlarged model training, pruning, and recovery under a single cosine annealing learning rate schedule. This system incorporates an iterative structured pruning technique to gradually eliminate parameters, redistributing model capacity across surviving neurons and avoiding the knowledge loss seen in naive approaches. Evaluated on compressing 2.8B parameter models to 1.3B across up to 2T pretraining tokens, the approach yields superior downstream model performance.


### [Luce: Relightable Gaussians for 3D Asset Generation](https://yomu.fyi/post/luce-relightable-gaussians-for-3d-asset-generation.md)
- Company: [Apple](https://yomu.fyi/company/apple.md)
- Published: Aug 26, 2026
- Type: Problem & solution

High-fidelity image-to-3D asset generation requires 3D representations that capture geometry alongside physically based rendering modalities for seamless pipeline integration and relighting. Luce addresses this challenge by unifying geometry and material properties into a voxelized multimodal Gaussian cloud with dedicated primitives for albedo, metallic-roughness, and surface normals. A variational autoencoder compresses these multimodal primitives into a unified material-aware latent space. A rectified-flow transformer subsequently generates this latent from a single input image using multi-layer features extracted from a pretrained image encoder. On the Toys4K dataset, Luce improves FID by 28% over the leading baseline while maintaining fine details such as text and logos.


### [Restore LLM Inference Capacity in Seconds with Shadow Engine Recovery in NVIDIA Dynamo](https://yomu.fyi/post/restore-llm-inference-capacity-in-seconds-with-shadow-engine-recovery.md)
- Company: [NVIDIA](https://yomu.fyi/company/nvidia-developer-blog.md)
- Author: Michelle Horton
- Published: Aug 25, 2026
- Type: Problem & solution

LLM engine process failures typically trigger cold restarts that reload weights into HBM, compile kernels, and recapture CUDA graphs, leaving surviving workers to absorb displaced traffic for minutes. Shadow engine recovery, a preview feature in NVIDIA Dynamo, keeps a fully initialized standby on the same GPUs as the active engine. GPU Memory Service (GMS) preserves one physical weight copy across engine processes, while the parked shadow precomputes its CUDA context, graphs, communicators, and mappings but defers KV cache materialization. In a GLM-5.2 test, the shadow resumed service in 7.3 seconds after a worker failure, compared with 283 seconds for a cold restart, reducing post-fault p50 TTFT from 23,815 ms to 1,311 ms and improving decode rate from 12 to 46 tok/s/user. The preview does not cover hardware, node, or multi-node failures and currently starts promoted shadows with empty KV caches.


### [How Capital Markets Finance Protects Balance Sheet Returns](https://yomu.fyi/post/how-capital-markets-finance-protects-balance-sheet-returns.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Andrea DeSosa, Kim Hatton
- Published: Aug 25, 2026
- Type: Problem & solution

Capital markets finance must protect returns on a balance sheet exposed continuously to market, funding, and liquidity risk, while funding costs, capital consumption, and liquidity needs can change intraday. Fragmented desk P&L, treasury, and risk data require shared business context, with ontology and governance linking numbers to desks, products, positions, funding, capital, and permissions. Databricks’ Genie One is presented as a data-smart AI coworker that lets authorized finance users investigate governed data, trace emerging funding or profitability issues, and surface the definitions and source evidence behind capital and valuation figures. One top global bank reduced regulatory data processing for liquidity reporting from 10 hours to 8 minutes, while pricing, funding, risk, and capital decisions remain with accountable people.


### [How Suprema Gaming made its data platform agent-ready with ClickHouse Cloud](https://yomu.fyi/post/how-suprema-gaming-made-its-data-platform-agent-ready-with-clickhouse.md)
- Company: [Clickhouse](https://yomu.fyi/company/clickhouse.md)
- Published: Aug 25, 2026
- Type: Problem & solution

Suprema Gaming migrated its analytics platform from Snowflake to ClickHouse Cloud on AWS as part of a company-wide move toward agentic operations. The previous stack delivered data about four hours late, required heavy dbt rebuilds, and often took minutes to answer queries, limiting real-time investigations and interactive analysis. Suprema rebuilt ingestion with ClickPipes for supported sources and Debezium plus Kafka for others, while dbt and GitHub Actions deploy the warehouse-layer structure without external cron jobs, Airflow, or a separate orchestrator; governed access uses role-based controls and query-time PII masking. Before cutover, every metric was validated against the reference source, and the resulting platform reduced warehouse spend by roughly 62%, improved direct-read performance 27x, and maintains about one-minute p95 freshness, supporting an early production deployment of ClickHouse Agents.


### [Project Lighthouse — Part 3: Introducing project-lighthouse-anonymize](https://yomu.fyi/post/project-lighthouse-part-3-introducing-project-lighthouse-anonymize.md)
- Company: [Airbnb](https://yomu.fyi/company/airbnb.md)
- Author: Adam Bloomston
- Published: Aug 25, 2026

Airbnb has open-sourced project-lighthouse-anonymize, a Python library powering their privacy-preserving anonymization used to measure user experience disparities. The release is accompanied by technical papers detailing Core Mondrian, an extended partition-based anonymization algorithm, alongside a framework for measuring anonymized data quality. The library enforces k-anonymity and p-sensitive k-anonymity while offering built-in data quality metrics.


### [Funding better evaluations of AI’s impact on wellbeing](https://yomu.fyi/post/funding-better-evaluations-of-ai-s-impact-on-wellbeing.md)
- Company: [Anthropic](https://yomu.fyi/company/anthropic.md)
- Published: Aug 25, 2026
- Type: Announcement

A $5 million grant program will fund independent research into how AI affects user wellbeing, offering direct funding, model access, and technical support for open-source evaluations. The initiative addresses the difficulty of judging wellbeing-related behavior, where risks may emerge only across long, shifting conversations and where otherwise reasonable advice can become harmful in a different context, such as disordered eating. Grantees will work independently and publish reusable evaluations or benchmarks, with guidance asking projects to define pass/fail criteria, involve clinical and subject-matter experts, test overcompliance and overrefusal, model realistic multi-turn scenarios, and validate graders against experts. Applications are due September 21, with selected full-proposal applicants notified by October 5.


### [Choosing Data Governance Tools for Enterprise Data Governance](https://yomu.fyi/post/choosing-data-governance-tools-for-enterprise-data-governance.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Databricks Staff
- Published: Aug 25, 2026
- Type: Explainer

Enterprise data governance tools address fragmented data estates in which teams struggle to locate assets, identify ownership, protect sensitive information, and maintain consistent quality across warehouses, lakes, SaaS applications, and spreadsheets. Positioned above storage and compute, these tools read metadata, schemas, and query logs to provide a shared governance layer for cataloging, lineage, classification, policy enforcement, quality monitoring, and audit reporting. The guide distinguishes standalone catalogs, point solutions, enterprise suites, platform-native tools, and open-source options while treating AI and agent governance as an increasingly important capability for models, prompts, and autonomous agents. It recommends evaluating candidates against seven criteria, comparing total cost of ownership, and planning a phased rollout that starts with highest-risk domains; a lakehouse-native approach can reduce synchronization between separate governance systems.


[Newer posts](https://yomu.fyi/posts/page/66.md) · [Older posts](https://yomu.fyi/posts/page/68.md)