---
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

### [ClickHouse welcomes RunReveal](https://yomu.fyi/post/clickhouse-welcomes-runreveal.md)
- Company: [Clickhouse](https://yomu.fyi/company/clickhouse.md)
- Published: Sep 1, 2026
- Type: Announcement

ClickHouse has acquired RunReveal, a company that built a security data platform on ClickHouse, and says the team’s expertise will become part of its work in security. The announcement frames security as a demanding data workload involving cloud audit logs, identity events, endpoint telemetry, and network flows, with continuous high-throughput ingest, years-long retention, and interactive query requirements. ClickHouse says the acquisition will bring sharper reference architectures, guidance on schema and retention design, and a clearer connection between customer requirements and its database roadmap, while RunReveal remains available through a bring-your-own-database model. Existing RunReveal customers will continue to receive platform support under unchanged contract terms, and the team’s work on agentic investigation is expected to inform support for analytics involving agents that hunt across sources, tune detections, and advance investigations.


### [The Agentic Analytics Benchmark: Measuring model accuracy and efficiency in analytical agents](https://yomu.fyi/post/the-agentic-analytics-benchmark-measuring-model-accuracy-and-efficienc.md)
- Company: [Clickhouse](https://yomu.fyi/company/clickhouse.md)
- Published: Sep 1, 2026
- Type: Benchmark

The post presents data-agent-mnist, an open harness for measuring analytical agents on correctness, cost, and speed against a warehouse-shaped workload rather than a single text-to-SQL query. The authors extracted 201 production analytics questions from DWAINE, curated and anonymized them, reconstructed a deterministic synthetic warehouse with 18 tables and 865 columns, and evaluated 28 proprietary and open-weight models. The benchmark preserves schema discovery, multi-turn querying, planning, and multi-hop joins across curated marts and a dimensional CRM layer, with provider-diverse ground truth and judging described as part of the workflow. On the authors’ warehouse, Claude Fable 5.1 ranked first for correctness at 76.6%, while the full run cost $1 with DeepSeek V4 Flash versus $52 with Fable 5.1, an 11-percentage-point correctness difference. The harness is intended to identify the best production fit for a specific warehouse, not a universal winner.


### [Collaboration makes us all stronger](https://yomu.fyi/post/collaboration-makes-us-all-stronger.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Aaron Kobayashi, Mehmet D. Ince, Anurag Srivastava, Alexey Kondratov
- Published: Sep 1, 2026
- Type: Incident / postmortem

An external researcher, Mehmet Ince, reported a memory-safety flaw in PostGIS’s address\_standardizer extension, where caller-controlled grammar data indexed a fixed-size array without bounds checking. Because the extension was available to ordinary tenant roles on managed Postgres services, Databricks treated the issue as a platform exposure rather than a third-party problem, while noting that its microVM architecture meant the exploit did not produce cross-customer impact on Lakebase Postgres and Neon. Detection alerted the Neon team during testing, enabling rapid validation and downstream patching while a durable fix was developed and deployed to protect tenants without customer action. The upstream fix initially missed cases; Mehmet supplied the remaining pieces, and the complete correction was returned to PostGIS, with his bounty donated and matched for the project.


### [How Claude's text watermarking works](https://yomu.fyi/post/how-claude-s-text-watermarking-works.md)
- Company: [Anthropic](https://yomu.fyi/company/anthropic.md)
- Published: Sep 1, 2026
- Type: Explainer

Anthropic describes a text-watermarking system planned for Claude to estimate whether Claude was involved in generating or processing text, primarily to comply with the EU AI Act. The method uses a key and preceding words to influence the source of randomness in low-stakes word choices, creating a detectable pattern without adding characters, tokens, identifying information, or extra cost. Anthropic says internal testing found no impact on content, creativity, readability, or perceived quality, while detection confidence improves with longer passages and weakens when text offers few valid alternatives, such as factual statements, proofreading, or code. The watermark can indicate only that Claude was likely involved, not whether text was human-written, who authored it, or which person or organization used Claude; light editing may not remove it, but a complete rewrite will.


### [New system views in PostgreSQL 19](https://yomu.fyi/post/new-system-views-in-postgresql-19.md)
- Company: [Clickhouse](https://yomu.fyi/company/clickhouse.md)
- Published: Sep 1, 2026
- Type: Explainer

PostgreSQL 19 adds four system views for lock contention, recovery, autovacuum priorities, and dynamic shared memory allocations. The article presents pg\_stat\_lock as cumulative, cluster-wide statistics with one row per lock type, reporting waits, total wait time, fastpath\_exceeded, and stats\_reset; waits and wait\_time count only successfully acquired locks that exceeded deadlock\_timeout, while fast-path overflow attempts are counted separately. Examples reset statistics, reproduce a 2.5-second table lock, and scan 140 partitions, producing 1 wait, 2201.783 milliseconds, and 422 fast-path overflows. The post describes recovery and autovacuum views, noting that autovacuum scores suggest likely priorities rather than guarantees because scoring and worker scheduling can differ. DSM registry entries expose name, type, and size; NULL means initialization failed.


### [Building an Adaptive Agentic Cybersecurity System with NVIDIA Nemotron](https://yomu.fyi/post/building-an-adaptive-agentic-cybersecurity-system-with-nvidia-nemotron.md)
- Company: [NVIDIA](https://yomu.fyi/company/nvidia-developer-blog.md)
- Author: Michelle Horton
- Published: Sep 1, 2026
- Type: Explainer

NVIDIA and CrowdStrike describe an agentic cybersecurity system that links red-agent attack execution with blue-agent detection engineering in a continuous offense-defense testing loop. In an isolated environment modeled on NVIDIA accelerated computing infrastructure, Falcon sensors captured telemetry while Nemotron 3 Ultra orchestrated the defensive workflow and a customized Nemotron 3 Super generated or repaired detections. The harness grounded agents in sensor schemas and attack traces, then applied linting, replay, structured correction, and independent review before deployment. Backtesting raised mean detection of the recorded attack from 16.5% to 41.9% with the optimized open pipeline across independently seeded sessions. In live-fire tests, five of 11 open detections detected an unseen attack, three qualified as gold, and those three covered all eight attacks; the authors call the result a directional case study because it used one scenario family and limited benign traffic.


### [How Uken Games reduces observability costs by 87% with ClickHouse](https://yomu.fyi/post/how-uken-games-reduces-observability-costs-by-87-with-clickhouse.md)
- Company: [Clickhouse](https://yomu.fyi/company/clickhouse.md)
- Published: Sep 1, 2026
- Type: Problem & solution

Uken Games replaced Datadog with an open-source observability stack built around ClickHouse to monitor backend services supporting mobile games with millions of players. The migration addressed rising Datadog costs and vendor lock-in while targeting pragmatic feature parity across metrics, traces, and logs. OpenTelemetry collectors run in two layers: lightweight agents on ECS instances pre-filter traces, while an autoscaling gateway batches and buffers writes to SigNoz on ClickHouse; managed Prometheus handles metrics, CloudWatch retains logs, and Grafana provides dashboards and alerts. Uken samples ordinary traces, retains errors and high-latency requests, computes alert metrics without sampling, and limits retention to two weeks. The resulting system stores all traces on one ClickHouse node using about 170 GB, operates without migration downtime, and costs about $12,000 annually—an 87% reduction.


### [Becoming an AI Team](https://yomu.fyi/post/becoming-an-ai-team.md)
- Company: [Pinterest](https://yomu.fyi/company/pinterest.md)
- Author: Pinterest Engineering
- Published: Sep 1, 2026
- Type: Explainer

Becoming an AI team requires more than adding assistants to existing workflows: it changes ownership, planning, roles, and execution. At Pinterest, infrastructure teams face the scale of serving billions of Pins, boards, ads, and real-time signals, making AI adoption an operational necessity for reliability, cost efficiency, and developer productivity. AI code-generation and transformation models can turn a widespread legacy-codebase refactor, such as replacing “foo” with “bar,” from a months-long effort into work completed and verified in a week or less. As routine execution is increasingly augmented or automated, engineers, product managers, and designers are expected to spend more time on strategy, problem definition, and user validation, while managers emphasize vision, trade-offs, mentorship, and collective impact. The conclusion is that teams must continuously optimize an AI-driven operating system rather than treat AI as a side project.


### [How to Size GPUs for AI Inference and TCO Without Overspending](https://yomu.fyi/post/how-to-size-gpus-for-ai-inference-and-tco-without-overspending.md)
- Company: [NVIDIA](https://yomu.fyi/company/nvidia-developer-blog.md)
- Author: Elizabeth Goodman
- Published: Sep 1, 2026
- Type: Tutorial

GPU sizing for AI inference depends on use case, token patterns, latency targets, concurrency, cache hit rate, model choice and deployment strategy rather than raw hardware specifications alone. The post proposes a core-and-flex capacity model, combining baseline on-prem or reserved cloud capacity with elastic spot or on-demand GPUs for surges, launches and experiments. It also recommends matching GPU memory to prompt lengths and concurrency, while using quantization, pruning and distillation to reduce footprint and operating cost. A NeMo example prunes Qwen3-8B from 36 to 24 layers or narrows its dimensions toward a roughly 6B student before teacher-student distillation. In the illustrative run, width pruning reached lower final validation loss than depth pruning, 3.21 versus 3.60, while depth pruning converged faster.


### [Operationalizing Genie Ontology in Your Data Stack](https://yomu.fyi/post/operationalizing-genie-ontology-in-your-data-stack.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Srujan Alase, Richard Tomlinson
- Published: Sep 1, 2026
- Type: Explainer

Genie Ontology is presented as a way to give enterprise AI agents shared business context beyond a semantic model, including definitions, relationships, business rules, authoritative sources, and permissions. It combines Unity Catalog Semantics—Metric Views, Pages, and Domains—with context inferred from governed tables, queries, dashboards, notebooks, and other supported assets. The guidance recommends six progressive layers, beginning with clean gold data and resolved golden records, then metadata, semantic modeling, enterprise context, governance, and evaluation. Critical implementation details include declaring informational primary and foreign keys, defining canonical measures in Metric Views, adding synonyms and example queries, and using permissions plus human-reviewed automation. Rather than waiting for complete coverage, it advises starting with one high-value domain and metric, then using feedback, telemetry, benchmarks, and drift reviews to strengthen trust over time.


### [Hybrid cloud orchestration: Modernizing on-premises infrastructure management with AWS](https://yomu.fyi/post/hybrid-cloud-orchestration-modernizing-on-premises-infrastructure-mana.md)
- Company: [AWS](https://yomu.fyi/company/aws.md)
- Author: Sandeep Singh
- Published: Sep 1, 2026
- Type: Explainer

Managing distributed on-premises infrastructure across hundreds of sites creates inconsistency, manual lifecycle bottlenecks, fragmented visibility, and scalability problems, especially when hardware, operating system, Kubernetes, and application operations span thousands of servers. The proposed design centralizes orchestration in AWS while keeping execution on site, using hybrid connectivity, AWS Lambda, AWS Step Functions, Amazon EventBridge, DynamoDB, AWS Systems Manager, and supporting services. Redfish APIs provide vendor-agnostic bare-metal control for BIOS, firmware, power, and health operations, while EKS Anywhere deploys and operates management and workload clusters on local hardware; orders and inventory records track workflows and resource relationships. DynamoDB Streams also automate DNS, certificates, AWS access, and storage integrations, while centralized observability aggregates server, cluster, and application signals; the post presents architectural patterns and defers detailed implementation workflows and templates to later parts.


### [The new Brickbuilder Partner Network tiers for ISVs and Data Providers are here](https://yomu.fyi/post/the-new-brickbuilder-partner-network-tiers-for-isvs-and-data-providers.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Stephen Orban
- Published: Sep 1, 2026
- Type: Announcement

The Databricks Partner Network has launched Bronze, Silver, Gold, and Platinum tiers for ISVs and data providers, with placement based on customer impact, technical excellence, strategic alignment, readiness and enablement, and joint go-to-market activity. Technical Excellence is evaluated through the Partner Well-Architected Framework (PWAF), which provides prescriptive, AI-ready guidance for secure, reliable integrations and serves as the standard for validated architecture. For customers, the tiers signal that a partner solution has met a high bar for architecture, security, compliance, and joint customer success, reducing vetting effort. For partners, each tier defines requirements and a roadmap tied to greater visibility, co-selling, and go-to-market support, while Databricks plans to refine the program and expand benefits over the coming months.


### [How Discovery Bank delivers hyper-personalized banking at scale: behavioral AI, governed data, and real-time decisioning](https://yomu.fyi/post/how-discovery-bank-delivers-hyper-personalized-banking-at-scale-behavi.md)
- Company: [Databricks](https://yomu.fyi/company/databricks.md)
- Author: Jack Yallop, Stuart Emslie
- Published: Sep 1, 2026
- Type: Explainer

Discovery Bank’s hyper-personalized banking model addresses how to make interactions relevant to individual behavior while meeting financial-services demands for scale, speed, security, and governance. The bank combines demographic, transactional, digital-engagement, savings, borrowing, credit-risk, rewards, and lifestyle-related data on the Databricks Data and AI Platform, producing governed reusable features, indicators, scores, forecasts, and recommendations. Behavioral models and a reusable next-best-action decisioning layer support personalization, fraud detection, servicing, and banker assistance; TRUST alerts assess transactions against client-specific norms and can escalate from explanation to account locking. A four-layer architecture adds control services and specialized generative AI and agents to trusted data and analytical foundations, with governance remaining in the loop. Reported results include a 40% uplift in client engagement impact, 20x faster pipeline development and data processing, 5x faster data-product creation, more than 300 models built per day, and return on investment above 500%.


### [From Chaos to Control: Addressing Shard Distribution Challenges in M3DB with Subclusters](https://yomu.fyi/post/from-chaos-to-control-addressing-shard-distribution-challenges-in-m3db.md)
- Company: [Uber](https://yomu.fyi/company/uber.md)
- Published: Sep 1, 2026
- Type: Problem & solution

M3DB’s sharded placement algorithm lets replicas spread across nodes subject primarily to isolation-group constraints, but large clusters experienced broad failure and maintenance impacts. The subclustered placement algorithm partitions nodes into fixed-size, self-contained groups, each owning a distinct shard slice while preserving replica separation across isolation groups. It requires equal instance weights, scaling in multiples of the subcluster size, a replica factor that cannot change through AddReplica, and instancesPerSubcluster divisible by RF. During scaling, a greedy donation strategy simulates removing each candidate shard, ranks the resulting per-node skew, and chooses the shard that leaves the donor most balanced, avoiding a second rebalancing move. The resulting model contains recovery within subclusters, enables parallel maintenance, and reduces arbitrary shard sharing from O(cluster) to O(subcluster), while the original algorithm remains suitable for smaller clusters or impractical scaling steps.


### [MCP went stateless: Is your AWS MCP server deployment well-architected?](https://yomu.fyi/post/mcp-went-stateless-is-your-aws-mcp-server-deployment-well-architected.md)
- Company: [AWS](https://yomu.fyi/company/aws.md)
- Author: Anand Komandooru
- Published: Sep 1, 2026
- Type: Problem & solution

MCP 2026-07-28 removes the initialize handshake and Mcp-Session-Id, making the protocol core stateless and changing how remote servers scale on AWS. Requests carry protocol version and client context, while server/discover can expose versions, capabilities, and identity; continuity remains possible through server-minted identifiers passed in tool arguments and stored application state. It covers round-robin routing, header-based gateway signals, cache freshness fields, W3C Trace Context, idempotent tools, and Multi Round-Trip Requests for input-dependent calls. It recommends retaining sticky routing and session stores for 2025-era clients, then migrating after legacy traffic reaches zero and validating against the official conformance suite. The conclusion is that the revision aligns remote MCP deployments with AWS Well-Architected practices and permits removal of infrastructure required only by the old protocol.


### [How we could save petabytes of cache storage with Zstandard and Pingora](https://yomu.fyi/post/how-we-could-save-petabytes-of-cache-storage-with-zstandard-and-pingor.md)
- Company: [Cloudflare](https://yomu.fyi/company/cloudflare.md)
- Author: Aashi Patel
- Published: Sep 1, 2026
- Type: Problem & solution

Cloudflare prototyped Cache Transcoding, a Pingora-based architecture that stores eligible cache responses as Zstandard (zstd) rather than their original representation. On a fill, the proxy compresses qualifying uncompressed text at zstd level 3, records the compressed representation and original content length, keeps it compressed on disk and between Tiered Cache layers, and decodes it on the client-facing path. Eligibility requires a 200 OK response, compressible text, no existing Content-Encoding, known Content-Length of at least 4 KiB, and excludes several request and binary-content cases. In initial tests, eligible assets shrank to roughly one-third of their original on-disk size, with a 2.834x ratio, while encoding cost 4.31 ns per byte and decoding 1.56 ns; the modeled CPU increase stayed within budget, though broader testing remains necessary.


### [The Grails Plugin Has a New Home: Apache Grails](https://yomu.fyi/post/the-grails-plugin-has-a-new-home-apache-grails.md)
- Company: [Jetbrains](https://yomu.fyi/company/jetbrains.md)
- Author: Andrey Belyaev
- Published: Sep 1, 2026
- Type: Announcement

The Grails plugin for IntelliJ IDEA, providing framework-aware coding assistance, GSP tooling, and GORM integration, has moved to the Apache Grails project. The Apache Software Foundation now owns its source code, while the Apache Grails Project Management Committee oversees development, releases, and compatibility rather than IntelliJ IDEA release cycles. The Grails team builds, signs, and publishes updates to JetBrains Marketplace under the Apache Grails vendor listing; build 262.0.0 supports IntelliJ IDEA 2026.2. Existing users do not need to migrate, and the plugin ID remains unchanged for now, but beginning with 2026.2 it is no longer bundled by default with IntelliJ IDEA Ultimate. Users who cannot find it in the IDE can install it from Marketplace, where JetBrains will continue featuring the plugin.


### [Application Metrics caught my broken size estimator](https://yomu.fyi/post/application-metrics-caught-my-broken-size-estimator.md)
- Company: [Sentry](https://yomu.fyi/company/sentry.md)
- Author: Kyle Tryon
- Published: Sep 1, 2026
- Type: Problem & solution

Cliparr’s video size estimator was intended to predict output size from target bitrate multiplied by clip duration, but real-world results could vary because its VBR encoder changes bitrate with content. Rather than manually testing a matrix of formats, codecs, and quality settings, the author extracted the browser-based export engine into Cliparr Convert, a standalone static converter, and instrumented it with anonymous Sentry Application Metrics. Completed conversions record actual and estimated bytes, their ratio and delta, plus attributes such as output format, quality, and estimator basis; counters also track conversion completion and the progressive web app installation funnel. The resulting dashboard exposed an estimate that was off by 83% and supports grouping real usage by conversion conditions, while sending only numeric metrics and not media, filenames, or URLs.


### [Ensuring Code Compliance in Public Sector Software Projects](https://yomu.fyi/post/ensuring-code-compliance-in-public-sector-software-projects.md)
- Company: [Jetbrains](https://yomu.fyi/company/jetbrains.md)
- Author: Kerry Beetge
- Published: Sep 1, 2026
- Type: Explainer

Public sector software must satisfy data protection, security, accessibility, procurement, supply-chain, and infrastructure requirements while protecting sensitive citizen data. Compliance failures can expose organizations to breaches, penalties, legal costs, contract disputes, and reputational damage. The brief organizes common risks around secure development, contractual and procurement obligations, audit evidence, long-term supportability, and IT governance, noting examples such as hardcoded credentials, weak cryptography, vulnerable dependencies, license conflicts, unsupported libraries, and unchecked quality issues. It recommends embedding automated SAST, secret detection, cryptography and dependency scans, license checks, quality gates, technical debt metrics, freshness tracking, testing, and time-stamped reports into IDE and CI/CD workflows, plus self-hosted or air-gapped deployment where policy requires it. These controls are presented as ways to catch problems before production, create audit trails, support handovers, and maintain compliance over long software lifecycles.


### [Orchestrate production RAG with OpenShift AI](https://yomu.fyi/post/orchestrate-production-rag-with-openshift-ai.md)
- Company: [Red Hat](https://yomu.fyi/company/red-hat.md)
- Author: Ana Biazetti, Saad Zaher
- Published: Sep 1, 2026
- Type: Tutorial

This guide shows how to move a streaming retrieval-augmented generation (RAG) workflow from a monolithic Ray Data script to modular AI pipelines on OpenShift AI. It addresses failure recovery and run traceability by separating parsing and chunking, ingestion, model download, and deployment into five reusable components. Parsed and chunked JSONL is stored in S3-compatible storage so KFP can rerun ingestion without repeating document processing, and each run records parameters, timestamps, logs, and status. The data chain uses Docling’s HybridChunker and Ray map\_batches with ActorPoolStrategy, while the model chain downloads a Hugging Face model and deploys it through vLLM and KServe in parallel. For embedding ingestion, the workflow supports local CPU processing with granite-embedding-125m-english or a dedicated GPU vLLM service for larger collections and offers notebooks for deployment and query validation.


[Newer posts](https://yomu.fyi/posts/page/59.md) · [Older posts](https://yomu.fyi/posts/page/61.md)