<!DOCTYPE html><html lang="en"><head><meta http-equiv="Content-Type" content="text/html charset=UTF-8"><meta charset="UTF-8"><meta name="viewport" content="width=device-width"><meta name="x-apple-disable-message-reformatting"><title>TLDR Data</title><meta name="color-scheme" content="light dark"><meta name="supported-color-schemes" content="light dark"><style type="text/css">
:root {
color-scheme: light dark; supported-color-schemes: light dark;
}
*,
*:after,
*:before {
-webkit-box-sizing: border-box; -moz-box-sizing: border-box; box-sizing: border-box;
}
* {
-ms-text-size-adjust: 100%; -webkit-text-size-adjust: 100%;
}
html,
body,
.document {
width: 100% !important; height: 100% !important; margin: 0; padding: 0;
}
body {
-webkit-font-smoothing: antialiased; -moz-osx-font-smoothing: grayscale; text-rendering: optimizeLegibility;
}
div[style*="margin: 16px 0"] {
margin: 0 !important;
}
table,
td {
mso-table-lspace: 0pt; mso-table-rspace: 0pt;
}
table {
border-spacing: 0; border-collapse: collapse; table-layout: fixed; margin: 0 auto;
}
img {
-ms-interpolation-mode: bicubic; max-width: 100%; border: 0;
}
*[x-apple-data-detectors] {
color: inherit !important; text-decoration: none !important;
}
.x-gmail-data-detectors,
.x-gmail-data-detectors *,
.aBn {
border-bottom: 0 !important; cursor: default !important;
}
.btn {
-webkit-transition: all 200ms ease; transition: all 200ms ease;
}
.btn:hover {
background-color: #f67575; border-color: #f67575;
}
* {
font-family: Arial, Helvetica, sans-serif; font-size: 18px;
}
@media screen and (max-width: 600px) {
.container {
width: 100%; margin: auto;
}
.stack {
display: block!important; width: 100%!important; max-width: 100%!important;
}
.btn {
display: block; width: 100%; text-align: center;
}
}
body,
p,
td,
tr,
.body,
table,
h1,
h2,
h3,
h4,
h5,
h6,
div,
span {
background-color: #FEFEFE !important; color: #010101 !important;
}
@media (prefers-color-scheme: dark) {
body,
p,
td,
tr,
.body,
table,
h1,
h2,
h3,
h4,
h5,
h6,
div,
span {
background-color: #27292D !important; color: #FEFEFE !important;
}
}
a {
color: inherit !important; text-decoration: underline !important;
}
</style><!--[if mso | ie]>
<style type="text/css">
a {
background-color: #FEFEFE !important; color: #010101 !important;
}
@media (prefers-color-scheme: dark) {
a {
background-color: #27292D !important; color: #FEFEFE !important;
}
}
</style>
<![endif]--></head><body class="">
<div style="display: none; max-height: 0px; overflow: hidden;">DocumentDB aims to keep the document model within an open, multi-vendor ecosystem, similar to how PostgreSQL itself thrived </div>
<div style="display: none; max-height: 0px; overflow: hidden;">
<br>
</div>
<table align="center" class="document"><tbody><tr><td valign="top">
<table align="center" border="0" cellpadding="0" cellspacing="0" class="container" width="600"><tbody><tr class="inner-body"><td>
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr class="header"><td bgcolor="" class="container">
<table width="100%"><tbody><tr><td class="container">
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" style="margin-top: 0px;" width="100%"><tbody><tr><td style="padding: 0px;">
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div style="text-align: center;">
<span style="margin-right: 0px;"><a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Ftldr.tech%2Fdata%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/-mv6gtL71uVNk1xWXli2tpxUKSkHK4oCBS8Lp_VbCrQ=421" rel="noopener noreferrer" target="_blank"><span>Sign Up</span></a>
|<span style="margin-right: 2px; margin-left: 2px;"><a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fadvertise.tldr.tech%3Futm_source=tldrdata%26utm_medium=newsletter%26utm_campaign=advertisetopnav/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/Dm0hQK6vyEwaSb5nnKvmlLVaWUoK_zHqRy5Ml1QRgy8=421" rel="noopener noreferrer" target="_blank"><span>Advertise</span></a></span>|<span style="margin-left: 2px;"><a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fa.tldrnewsletter.com%2Fweb-version%3Fep=1%26lc=1670a604-84b7-11f0-bcf5-55fc1d40139c%26p=78838578-893f-11f0-a24d-81099bb8a2b6%26pt=campaign%26t=1756980343%26s=bc9f020b624c9ba34dcede70a9c84a4feafca7b9d553887b3513b52ddac81d5d/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/15YuizYn5gF0ykGSBzD3pa5z7sWlU7IEXYCjUMxNbd0=421"><span>View Online</span></a></span>
<br>
</span></div>
</td></tr></tbody></table>
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="text-align: center;"><span data-darkreader-inline-color="" style="--darkreader-inline-color:#3db3ff; color: rgb(51, 175, 255) !important; font-size: 30px;">T</span><span style="font-size: 30px;"><span data-darkreader-inline-color="" style="color: rgb(232, 192, 96) !important; --darkreader-inline-color:#e8c163; font-size:30px;">L</span><span data-darkreader-inline-color="" style="color: rgb(101, 195, 173) !important; --darkreader-inline-color:#6ec7b2; font-size:30px;">D</span></span><span data-darkreader-inline-color="" style="--darkreader-inline-color:#dd6e6e; color: rgb(220, 107, 107) !important; font-size: 30px;">R</span>
<br>
</td></tr></tbody></table>
<br>
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr id="together-with"><td align="center" height="20" style="vertical-align:middle !important;" valign="middle" width="100%"><strong style="vertical-align:middle !important; height: 100%;">Together With </strong>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fsummit.starrocks.io%2F2025%2FTLDR/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/kIJPRBKf0rimu1m0yT0mII-Z0Gutpqz8LEgiO-u6gOU=421"><img src="https://images.tldr.tech/celerdata.png" valign="middle" style="vertical-align: middle !important; height: 100%;" alt="CelerData"></a></td></tr></tbody></table>
<table style="table-layout: fixed; width:100%;" width="100%"><tbody><tr><td style="padding:0;border-collapse:collapse;border-spacing:0;margin:0;">
<div style="text-align: center;">
<h1><strong>TLDR Data <span id="date">2025-09-04</span></strong></h1>
</div>
</td></tr></tbody></table>
<table style="table-layout: fixed; width:100%;" width="100%"><tbody><tr id="sponsy-copy"><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fsummit.starrocks.io%2F2025%2FTLDR/2/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/SOMKS6gqVAH7TAcWWsWLpEg9v7WfqF-xzzXGcrC3YzM=421">
<span>
<strong>49 clusters → 1 cluster: learn how to scale results without scaling resources at StarRocks Summit (Sponsor)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Most data teams are overwhelmed with pipelines, dashboards, and feature requests. But at <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fsummit.starrocks.io%2F2025%2FTLDR/3/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/bP4ybLWbSAIR3hFcZPkkuORO8mPApJgQKYMx0xkk02Y=421" rel="noopener noreferrer nofollow" target="_blank"><span>StarRocks Summit 2025 (free, virtual)</span></a>, you'll hear from teams who are managing to deliver rapid wins for their companies.
<p></p>
<p>You'll get a chance to <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fsummit.starrocks.io%2F2025%2FTLDR/4/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/lnoXYOu_t8Hz5luQXEhaMhcEPm1rcDF6Rw3XLfM50ZQ=421" rel="noopener noreferrer nofollow" target="_blank"><span>meet data teams at Coinbase, Intuit, Pinterest, and Demandbase,</span></a> and learn the details of real-life success stories such as:</p>
<p>→ Collapsing 49 clusters into 1... and getting FASTER queries</p>
<p>→ Cutting query planning from 65s to 6s with StarRocks + Apache Iceberg</p>
<p>→ Savings 90% on storage costs</p>
<p>This is an engineer-to-engineer conference with no sales decks allowed. If that sounds like your thing, <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fsummit.starrocks.io%2F2025%2FTLDR/5/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/VlFS0zuXraCn5HYvR2TD_dUZLjxgjzbmrtnqnGfLVfI=421" rel="noopener noreferrer nofollow" target="_blank"><span>save your spot on September 10</span></a>
</p>
</span></span></div>
</td></tr></tbody></table>
</td></tr></tbody></table>
</td></tr></tbody></table>
</td></tr>
<tr bgcolor=""><td class="container">
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td style="padding: 0px;">
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding-top: 0px; padding-bottom: 0px;">
<div class="text-block">
<div style="text-align: center;"><span style="font-size: 36px;">📱</span></div></div>
</td></tr></tbody></table>
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding-top: 0px; padding-bottom: 0px;">
<div class="text-block">
<div style="text-align: center;">
<h1><strong>Deep Dives</strong></h1>
</div>
</div>
</td></tr></tbody></table>
<table style="table-layout: fixed; width: 100%;" width="100%"><tbody><tr><td style="padding:0;border-collapse:collapse;border-spacing:0;margin:0;" valign="top">
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fjack-vanlightly.com%2Fblog%2F2025%2F9%2F2%2Funderstanding-apache-fluss%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/9AxJYBS6xhIj8PPhmXAGLFmmbWSxLzbbm87CDyZHNu8=421">
<span>
<strong>Understanding Apache Fluss (26 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Apache Fluss is an emerging low-latency disaggregated table storage engine purpose-built for Apache Flink that offers both append-only and primary key tables with efficient changelog capabilities. This addresses the latency and changelog limitations of Apache Paimon for real-time analytics. Fluss features a cluster of tablet and coordinator servers and columnar Arrow IPC storage with projection pushdown. It integrates seamlessly with object stores and lakehouse formats like Paimon. Its hybrid architecture enables unified, efficient access to both hot and historical data, empowering fast lookup joins and offloading Flink job state to shared tables.
</span>
</span>
</div>
</td></tr></tbody></table>
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fmedium.com%2Ffresha-data-engineering%2Ficeberg-mor-the-hard-way-starrocks-code-dive-fee5e1be66f5%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/qKpxT0GLEKCspS8Nw_U-9Kiz2bts1QOECOYV3M5rMG4=421">
<span>
<strong>Iceberg MoR the Hard Way: StarRocks Code Dive (13 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
StarRocks simplifies Iceberg's complex Merge-on-Read semantics by cleanly splitting responsibilities: the frontend handles metadata, routing, and MoR logic, while the backend just runs simple primitives like bitmap scans for positional deletes and anti-joins for equality deletes. Queues and incremental processing ensure parallelism and bounded memory. This MoR design reduces runtime flexibility (vs. engines like Spark or Trino) in exchange for predictable, efficient execution.
</span>
</span>
</div>
</td></tr></tbody></table>
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fwww.alibabacloud.com%2Fblog%2Fflink-state-management-a-journey-from-core-primitives-to-next-generation-incremental-computation_602503%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/uyDJX5HNFdgYNit2zJ3MhnNpiAYiGWhNGtaydg4GaUw=421">
<span>
<strong>Flink State Management: A Journey from Core Primitives to Next-Generation Incremental Computation (15 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Apache Flink's state management has evolved from embedded local state in Flink 1.x, using JVM heap and local disks, to a cloud-native disaggregated architecture in Flink 2.0 with the ForSt backend, enabling unlimited state capacity, zero-copy operations, instant recovery, and efficient resource utilization. Future generic incremental computation aims to unify stream and batch processing for cost-effective, real-time analytics.
</span>
</span>
</div>
</td></tr></tbody></table>
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fwww.wix.engineering%2Fpost%2Fhow-wix-cut-50-of-its-data-platform-costs-without-sacrificing-performance-part-2%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/GwM7RToahV6XrL1tFpOcyqZbr1t7Z3EUh3ia3lgpksk=421">
<span>
<strong>How Wix Cut 50% of Its Data Platform Costs - Without Sacrificing Performance (6 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Wix cut its data platform costs by 50% without sacrificing performance by implementing job-level telemetry, assigning clear ownership to data lake schemas, and creating a simple pricing model based on storage and CPU usage. It also introduced efficiency KPIs, built actionable Grafana dashboards, and fostered ongoing cost optimization through team engagement and infrastructure improvements like data compaction and tiered storage.
</span>
</span>
</div>
</td></tr></tbody></table>
</td></tr></tbody></table>
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding-top: 0px; padding-bottom: 0px;">
<div class="text-block">
<div style="text-align: center;"><span style="font-size: 36px;">🚀</span></div>
</div>
</td></tr></tbody></table>
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding-top: 0px; padding-bottom: 0px;">
<div class="text-block">
<div style="text-align: center;">
<h1><strong>Opinions & Advice</strong></h1>
</div>
</div>
</td></tr></tbody></table>
<table style="table-layout: fixed; width: 100%;" width="100%"><tbody><tr><td style="padding:0;border-collapse:collapse;border-spacing:0;margin:0;" valign="top">
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fmedium.com%2F@AnalyticsAtMeta%2Fdata-scientists-framework-for-navigating-product-strategy-as-data-leaders-2eb62b20f505%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/iDhfMSJG1wqLPFyRTlL_xQSTY4S67CBKUzwY1X_Enaw=421">
<span>
<strong>Meta's Data Scientist's Framework for Navigating Product Strategy as Data Leaders (7 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Meta's Data Scientist's Framework guides data scientists in shaping product strategy across four scenarios: Pioneer (low data, broad problem), Craftsperson (low data, concrete problem), Explorer (high data, broad problem), and Optimizer (high data, concrete problem). The framework uses a tailored approach that involves setting North Star metrics, designing data collection, identifying patterns, and optimizing with continuous learning.
</span>
</span>
</div>
</td></tr></tbody></table>
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fsqlpatterns.com%2Fp%2Fwhat-ive-learned-designing-data-systems%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/l2V2ovSwxfQ5yG1Kv_eL6a3RmIKYe-bVvUDygIQT2rw=421">
<span>
<strong>What I've Learned Designing Data Systems (4 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Modular data system design remains crucial as evolving business requirements demand flexibility, with modular, self-contained components enabling independent development and easier maintenance. Adopting functional architectures significantly reduces operational pain and shifts focus from compute/storage cost to engineering time. Design the data model upfront and set on a modeling strategy that aligns with specific analytic goals.
</span>
</span>
</div>
</td></tr></tbody></table>
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fsemyonsinchenko.github.io%2Fssinchenko%2Fpost%2Fspark-and-jmh%2F%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/hRbWkoew3LzesgTwAc8sYP0AGU7ws0RGbW8i25NiZ2w=421">
<span>
<strong>Benchmarking Spark Library with JMH (8 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Spark's memory allocation via spark-submit is incompatible with JMH. Leverage JMH's feature to fork a new JVM with custom Java options like -Xmx10g to address OutOfMemory errors during GraphFrames benchmarking in an existing JVM. This will ensure sufficient memory allocation.
</span>
</span>
</div>
</td></tr></tbody></table>
</td></tr></tbody></table>
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding-top: 0px; padding-bottom: 0px;">
<div class="text-block">
<div style="text-align: center;"><span style="font-size: 36px;">💻</span></div>
</div>
</td></tr></tbody></table>
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding-top: 0px; padding-bottom: 0px;">
<div class="text-block">
<div style="text-align: center;">
<h1><strong>Launches & Tools</strong></h1>
</div>
</div>
</td></tr></tbody></table>
<table style="table-layout: fixed; width: 100%;" width="100%"><tbody><tr><td style="padding:0;border-collapse:collapse;border-spacing:0;margin:0;" valign="top">
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fredmonk.com%2Fsogrady%2F2025%2F09%2F02%2Fdocumentdb%2F%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/A5f6TCcNuPnhag-lEpJUvIwo5LrMvRXx8hdfg3MKe7E=421">
<span>
<strong>DocumentDB and the Future of Open Source (6 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
The Linux Foundation has taken stewardship of DocumentDB, a MongoDB-compatible database built on PostgreSQL. Unlike MongoDB's proprietary path, DocumentDB aims to keep the document model within an open, multi-vendor ecosystem, similar to how PostgreSQL itself thrived. This could be a pivotal moment for document workloads, offering open-standard alternatives that reduce lock-in while leveraging PostgreSQL's reliability and tooling.
</span>
</span>
</div>
</td></tr></tbody></table>
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fgithub.com%2Fdrawdb-io%2Fdrawdb%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/CmEqL7NqATd_NFGHP4O4CuQOcofTKFp0g6x-K8kM4p8=421">
<span>
<strong>drawDB (GitHub Repo)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
drawDB is a browser-based tool to design and visualize database schemas with ease. You can build entity‑relationship charts, export DDLs, and share or embed diagrams. It's open-source, supports specific MySQL, PostgreSQL, SQLite, and SQL Server syntaxes, and works offline.
</span>
</span>
</div>
</td></tr></tbody></table>
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fgithub.com%2Fmatsonj%2Fcsv-everything%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/TlvfbI0xtYpL7f30Z0avvr4ETQRfNexhiD0r0ipKQXU=421">
<span>
<strong>CSV Everything (GitHub Repo)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
CSV Everything is a Chrome Extension that converts images of tables or charts into downloadable CSV files using the OpenRouter API. It supports clipboard image input, customizable AI models and prompts, and integrates natively with Chrome's download system.
</span>
</span>
</div>
</td></tr></tbody></table>
</td></tr></tbody></table>
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding-top: 0px; padding-bottom: 0px;">
<div class="text-block">
<div style="text-align: center;"><span style="font-size: 36px;">🎁</span></div></div>
</td></tr></tbody></table>
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding-top: 0px; padding-bottom: 0px;">
<div class="text-block">
<div style="text-align: center;"><strong><h1>Miscellaneous</h1></strong></div>
</div>
</td></tr></tbody></table>
<table bgcolor="" style="table-layout: fixed; width: 100%;" width="100%"><tbody><tr><td style="padding:0;border-collapse:collapse;border-spacing:0;margin:0;" valign="top">
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fwww.pedronasc.com%2Farticles%2Flessons-building-ai-data-analyst%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/mjCoMrp2D5qvA4HdwaC8u13GJRwTbI38kvzK1vhtwAc=421">
<span>
<strong>Lessons on Building an AI Data Analyst (16 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Building an effective AI data analyst requires more than just text-to-SQL capabilities. It involves creating a multi-step analysis workflow that integrates context and external tools. A robust semantic layer enhances data understanding and reduces SQL complexity, while a multi-agent system optimizes retrieval and reasoning to deliver precise, actionable insights. Continuous evaluation and performance tuning are essential to meet user expectations for quality and responsiveness.
</span>
</span>
</div>
</td></tr></tbody></table>
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fengineering.fb.com%2F2025%2F09%2F02%2Fml-applications%2Fa-new-ranking-framework-for-better-notification-quality-on-instagram%2F%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/eWJ5Ki4hFebzS_AoFJIeKkfVZZNlcCoKdap5zIjpMrQ=421">
<span>
<strong>A New Ranking Framework for Better Notification Quality on Instagram (4 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Instagram has implemented a diversity-aware notification ranking framework that overlays a multiplicative penalty on traditional ML engagement scores, factoring in content, author, type, and product surface similarity. This approach leverages Maximal Marginal Relevance to dynamically downrank repetitive notifications, effectively reducing daily notification volume while boosting click-through rates.
</span>
</span>
</div>
</td></tr></tbody></table>
</td></tr></tbody></table>
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding-top: 0px; padding-bottom: 0px;">
<div class="text-block">
<div style="text-align: center;"><span style="font-size: 36px;">⚡</span></div></div>
</td></tr></tbody></table>
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding-top: 0px; padding-bottom: 0px;">
<div class="text-block">
<div style="text-align: center;">
<h1><strong>Quick Links</strong></h1>
</div>
</div>
</td></tr></tbody></table>
<table bgcolor="" style="table-layout: fixed; width: 100%;" width="100%"><tbody><tr><td style="padding:0;border-collapse:collapse;border-spacing:0;margin:0;" valign="top">
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fwww.cybertec-postgresql.com%2Fen%2Fpartitioned-table-statistics%2F%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/Qzr_iEt5RBJoA_ekuoPt9t1en7PntyFM66XItmQs3_g=421">
<span>
<strong>Partitioned Table Statistics (3 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
PostgreSQL's autovacuum does not collect optimizer statistics for partitioned tables, leading to inaccurate query estimates, especially for joins.
</span>
</span>
</div>
</td></tr></tbody></table>
<table align="center" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block">
<span>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fwww.fivetran.com%2Fpress%2Ffivetran-acquires-tobiko-data-to-power-the-next-generation-of-advanced-ai-ready-data-transformation%3Futm_source=tldrdata/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/x8k3oxIVvvCk_etI-ewLF99Dqq7d318jr1GsfUkJu3M=421">
<span>
<strong>Fivetran Acquires Tobiko Data to Power the Next Generation of Advanced, AI-Ready Data Transformation (4 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Fivetran has acquired Tobiko Data, the team behind SQLMesh and SQLGlot, to expand its end-to-end data platform with advanced transformation capabilities.
</span>
</span>
</div>
</td></tr></tbody></table>
</td></tr></tbody></table>
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td align="left" style="word-break: break-word; vertical-align: top; padding: 5px 10px;">
<p style="padding: 0; margin: 0; font-size: 22px; color: #000000; line-height: 1.6; font-weight: bold;">
Want to advertise in TLDR? 📰
</p>
<div class="text-block" style="margin-top: 10px;">
If your company is interested in reaching an audience of data engineering professionals and decision makers, you may want to <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fadvertise.tldr.tech%2F%3Futm_source=tldrdata%26utm_medium=newsletter%26utm_campaign=advertisecta/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/jJyv6TvAihkJ07BFkyb4Wf8tBm_3LTMyXOsLsAp2GuA=421"><strong><span>advertise with us</span></strong></a>.
</div>
<br>
<!-- New "Want to work at TLDR?" section -->
<p style="padding: 0; margin: 0; font-size: 22px; color: #000000; line-height: 1.6; font-weight: bold;">
Want to work at TLDR? 💼
</p>
<div class="text-block" style="margin-top: 10px;">
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fjobs.ashbyhq.com%2Ftldr.tech/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/kzdGxbDYhuwyD1xhV6E0NxIjsWSggSoMFPmfIlqTjq0=421" rel="noopener noreferrer" style="color: #0000EE; text-decoration: underline;" target="_blank"><strong>Apply here</strong></a> or send a friend's resume to <a href="mailto:jobs@tldr.tech" style="color: #0000EE; text-decoration: underline;">jobs@tldr.tech</a> and get $1k if we hire them!
</div>
<br>
<div class="text-block">
If you have any comments or feedback, just respond to this email!
<br>
<br> Thanks for reading,
<br>
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fwww.linkedin.com%2Fin%2Fjoelvanveluwen%2F/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/vkVO1xjTHfMF4uniRJZoor8n87Z1oBB7imKc45bCRoY=421"><span>Joel Van Veluwen</span></a>, <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fwww.linkedin.com%2Fin%2Fjennytzurueyching%2F/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/KyHzngjEm4Eut-PdfOO8iiGCNOsHLTwmjpSJRIwqDJ0=421"><span>Tzu-Ruey Ching</span></a> & <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fwww.linkedin.com%2Fin%2Fremi-turpaud%2F/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/DhtQ38xTned0ElQOaeAw1wZZbIgrCTfFuQ1v6YUSLD8=421"><span>Remi Turpaud</span></a>
<br>
<br>
</div>
<br>
</td></tr></tbody></table>
<table align="center" bgcolor="" border="0" cellpadding="0" cellspacing="0" width="100%"><tbody><tr><td class="container" style="padding: 15px 15px;">
<div class="text-block" id="testing-id">
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Ftldr.tech%2Fdata%2Fmanage%3Femail=silk.theater.56%2540fwdnl.com/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/rZ4jWKbeRqzfyBSR-XTH-dguTH1itY_xf6MAbATUews=421">Manage your subscriptions</a> to our other newsletters on tech, startups, and programming. Or if TLDR Data isn't for you, please <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fa.tldrnewsletter.com%2Funsubscribe%3Fep=1%26l=037ede50-92cc-11ee-b0f2-b761aa2217ad%26lc=1670a604-84b7-11f0-bcf5-55fc1d40139c%26p=78838578-893f-11f0-a24d-81099bb8a2b6%26pt=campaign%26pv=4%26spa=1756980052%26t=1756980343%26s=9f35772e741fa03e72c1931e16bf4455a9aacaba7ea0b4ba653b802f1f7e801a/1/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/LHOdmy3hS_jjsCyCypN9GOkYLazM-6N-5Dm6p0_AyvI=421">unsubscribe</a>.
<br>
</div>
</td></tr></tbody></table>
</td></tr></tbody></table>
</td></tr></tbody></table>
</td></tr></tbody></table>
</td></tr></tbody></table>
<img alt="" src="http://tracking.tldrnewsletter.com/CI0/010001991430739c-4db6cc23-9370-4ebd-8563-bcb135095430-000000/46lA2aUVqmulcyrt_HzpX8_gJoil3JqBtighbnZ7v-o=421" style="display: none; width: 1px; height: 1px;">
</body></html>