<!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;">AI coding is shifting from token maxing to token efficiency as teams move from subscriptions to per-token billing and costs become harder to control </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/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/WZSK8DTkUvm-3YtQ_3tFfF8MxP-qWUIhL7CmTxRSnzo=452" 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/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/Pe6fZUjzPg0wk2IZRe7HP0xRzbQ6jYuiD2Xp6lMZ9cA=452" 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=48058a66-736a-11f1-98dc-afc99fb97726%26pt=campaign%26t=1782727672%26s=003ca45a17b1268c7fcb6a492b6c34e76268e4fdc39ac428b062383a06bb5002/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/S3H6wnqBfDy8wxsv-B3myEaIqWrUl2W_HZx8MH2E20Y=452"><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></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">2026-06-29</span></strong></h1>
</div>
</td></tr></tbody></table>
<table style="table-layout: fixed; width:100%;" width="100%"><tbody></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%2Flinks.tldrnewsletter.com%2F2Cg04g/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/IT50KNwApABmOOjPhiAOh_gpEj3JRDihcR27u9eIOBs=452">
<span>
<strong>12TB of AI Coding Agent Logs (17 minute video)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
AI coding is shifting from token maxing to token efficiency as teams move from subscriptions to per-token billing and costs become harder to control. Better workflows rely on careful upfront planning, right-sized agent sessions, cleaner context, API-first tooling, strong CI, and focused human review.
</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%2Flinks.tldrnewsletter.com%2FOzJtfD/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/KLxIqSa0ul57Snk8PSpAIZsKpKtqSBTef_BI00bbP_E=452">
<span>
<strong>Automated Schema Evolution in Pinterest's Next-Generation DB Ingestion Framework (11 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Pinterest built schema evolution for CDC across Kafka, Flink, Spark, and Iceberg, treating schema as a contract. Source schemas and sink mappings generate Flink/Spark/Iceberg artifacts, while push- and pull-based checks detect drift. Changes roll out with PR auditability, SLA-based recovery, and backfill fallbacks.
</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.langchain.com%2Fblog%2Ffull-text-search-in-smithdb-constructing-and-querying-our-inverted-index-pt-2%3Futm_source=tldrdata/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/hqOehOS4ClAdwQn172pG17V5ZOD84qS4GmJ5n7nm1So=452">
<span>
<strong>How we built SmithDB's inverted index for full-text search (11 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
SmithDB builds inverted indexes with efficient JSON parsing, tokenization, string interning, and radix sorting; interning lifted construction speed by ~2.2x. Streaming compaction bounds memory regardless of index size, while aligned chunks and request coalescing reduce object-storage GETs. Queries merge local-SSD indexes with object-storage segments for sub-second freshness.
</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.razorpay.com%2Fturning-scattered-data-into-queryable-segments-at-scale-how-razorpay-built-its-customer-data-3937c4b012de%3Futm_source=tldrdata/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/r3pFhjucyVA1yiet38LkOgJeTW1XYJNeqFVoY3f58yg=452">
<span>
<strong>Turning Scattered Data Into Queryable Segments at Scale: How Razorpay Built Its Customer Data Platform (11 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Razorpay built an in-house Customer Data Platform to turn scattered transaction data across 500M+ user profiles into real-time, queryable audience segments, using Airflow DAGs + Spark for daily segment computation (with reuse and deduplication), Temporal workflows for reliable DynamoDB ingestion with zero-downtime versioning, and privacy-preserving hashed lookups.
</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%2Fwww.snowflake.com%2Fen%2Fblog%2Fengineering%2Fmeasuring-real-workload-performance%2F%3Futm_source=tldrdata/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/uIDIDrVRCxIWCdrGGElFbCZBqw8n3xkNPJS9zPuNFog=452">
<span>
<strong>Why Real Workload Performance is the Metric that Matters (7 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Real workload performance matters more than headline benchmarks because production systems need to handle real data, concurrency, latency, scale, and cost. Performance claims should be judged by whether the workload matches yours, the setup is production-ready, results hold as data grows, and the product is actually available.
</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%2Flinks.tldrnewsletter.com%2FbRndrZ/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/0Dvw3cLqWcTyScSs7yww8gkUE61NDW2e8JNGbGlpjM8=452">
<span>
<strong>Building My Own Self-Hosted dbt Cloud (6 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
A self-hosted dbt Cloud-style app can deliver much of the developer experience by combining dbt Core with a React/FastAPI interface and Prefect for orchestration. The biggest lesson is to use APIs, not CLI scraping, for reliable job management, logs, deployments, and real-time run status.
</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%2Fgo.predicthq.com%2Ftldr%2Bdata%2Bpredicthq%3Futm_source=tldrdata/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/stOfdWtQKcqJGAPhvmVb846kCjNsTNJ6nJ-lbBlqdMw=452">
<span>
<strong>AI lacks real-world context - and it's costing business <em>trillions</em> annually (Sponsor)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Even the best AI still misses 60% of demand variability - and the problem isn't the models. <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fgo.predicthq.com%2Ftldr%2Bdata%2Bpredicthq/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/W24qoimxBdBhrhR6IXm3PDYiOmWHXq8rTw_4lmlrjPI=452" rel="noopener noreferrer nofollow" target="_blank"><span>This PredictHQ guide</span></a> lays out how <strong>Uber</strong> and <strong>Domino's</strong> are grounding AI with real-world context from global events and demand data.<strong> </strong>Want to see the impact? <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fgo.predicthq.com%2Ftldr%2Bphq%2Bsignup/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/AFV8LilZEHFcccMf-RYqLTl27NFMMJN2dfkIW8oQd34=452" rel="noopener noreferrer nofollow" target="_blank"><span>Try the PredictHQ API and MCP free for 14 days.</span></a>
</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%2Fflink.apache.org%2F2026%2F06%2F25%2Fapache-flink-2.3.0-release-announcement%2F%3Futm_source=tldrdata/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/dbEqpWxlep5FHJDN0xDtLTI_IU9DD0ph3g-AQEaP3rE=452">
<span>
<strong>Apache Flink 2.3.0 Release Announcement (8 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Flink 2.3 moves toward a declarative streaming data platform. Materialized tables can evolve through DDL and query changes while avoiding unnecessary historical reprocessing in many common cases. SQL adds changelog conversion, explicit upsert conflict handling, and native S3 support without Hadoop dependencies.
</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.morling.dev%2Fblog%2Fhardwood-1-0-fast-lightweight-apache-parquet-reader-for-the-jvm%2F%3Futm_source=tldrdata/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/Oj-90tk3TycjtGECRwNZ3Z5i3rCLQCU1244vYxjQKCg=452">
<span>
<strong>Hardwood 1.0: A Fast, Lightweight Apache Parquet Reader for the JVM (9 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Hardwood 1.0 is a production-ready, JVM-native Parquet reader for Java 21+ that removes mandatory dependencies and parallelizes page decoding across CPU cores by default. It covers Parquet physical/logical types, projections, predicate push-down, local and object-store files, with row and batch column APIs. Benchmarks show 16.5M rows/sec and ~17-18x selective push-down speedups.
</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%2Fjack-vanlightly.com%2Fblog%2F2026%2F6%2F24%2Fkafka-share-groups-pathological-fetch-waits-with-recordlimit%3Futm_source=tldrdata/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/iJ-aU9lqgj2YED3tR6E0dhf-qpitbQ4-6WINL34LB8s=452">
<span>
<strong>Kafka Share Groups - Pathological Fetch Waits with Record_limit (13 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
A notable performance pitfall in Kafka Share Groups arises when using record_limit with fewer consumers than partitions, especially under partition skew. This leads to pathological fetch waits, which can drastically slow consumption during backlog drains or skewed workloads. The simplest mitigation is to use at least as many consumers as partitions when running with record_limit.
</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%2Fmanticoresearch.com%2Fblog%2Fonnx-embeddings-speedup%2F%3Futm_source=tldrdata/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/yvHGaZ029iUvxFznQ0L21bbkRjIrHM2yXQdCknQxr2o=452">
<span>
<strong>14x faster embeddings: how we rebuilt the ONNX path in Manticore (9 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Manticore rewrote its embedding pipeline on ONNX Runtime, slashing CPU waste and lifting throughput up to 14x for low-latency vector search. The design shares one thread-safe ONNX session, disables intra-op spinning, and processes documents individually to avoid lock contention and variable-length padding overhead.
</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%2Fwe-ran-250-ai-agent-evals-to-find-out-if-skills-beat-docs-the-answer-is-more-complicated-than-we-ex%3Futm_source=tldrdata/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/fygQ_pmXWQBZrhqCZSVBEGqXEn0THr2q-G9UGJk0TsM=452">
<span>
<strong>We Ran 250 AI Agent Evals to Find Out if Skills Beat Docs. The Answer Is More Complicated Than We Expected (6 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Wix's 250-run evaluation found agent-optimized docs improved CLI task completion from 67% to 87%, cut token use 35%, and beat skills-only runs when skills were stale or misaligned. For API tasks, both hit 80% completion, but docs ran 31% faster while skills used 29% fewer tokens. Use optimized docs as the foundation, with skills as an evaluated caching layer.
</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%2Fdropbox.tech%2Fmachine-learning%2Fhow-we-turned-ai-evaluations-into-better-responses-in-dash-chat%3Futm_source=tldrdata/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/dONQT6VZRaPJtK42_7-PJMkonzRAIo_yKatFY_jU_SU=452">
<span>
<strong>How we used DSPy to turn AI evaluations into better responses in Dash chat (5 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Dropbox used DSPy to turn AI evaluations into concrete Dash Chat improvements, combining LLM-as-judge evals, human-labeled examples, offline replay, and statistical validation. The result was fewer incomplete answers, better coverage of user intent, and lower token use without compromising answer quality.
</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%2Fcommoncrawl.org%2Fblog%2Fhost--and-domain-level-web-graphs-april-may-and-june-2026%3Futm_source=tldrdata/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/ZQGnbSEuxaFXQupngxW4XY1J5oknN9OiAZpmSAY8p9U=452">
<span>
<strong>Host- and Domain-Level Web Graphs April, May, and June 2026 (3 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Fresh host/domain graphs offer 6B+ edges for large-scale link analysis without running your own crawler.
</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%2Fhuggingface.co%2Fspaces%2Fgemma-challenge%2Fgemma-interactions-view%3Futm_source=tldrdata/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/DjfXdTqrrhNXrAUUauuqCc1k6R6PMsUpgAxnrzMd5g0=452">
<span>
<strong>Gemma Interactions View (5 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
A coding-agent challenge turned into a collaborative lab, with agents sharing playbooks, pooling quota, debugging each other's work, and stacking small improvements into big performance gains.
</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/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/44kilrCilVtAZDnI5q6kpwO5w5YaicKrhNY0_jU7zOM=452"><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/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/JcouzvHOQPt8KJLDge6nDVF4ykkohFVvS5PXCGcAKhk=452" rel="noopener noreferrer" style="color: #0000EE; text-decoration: underline;" target="_blank"><strong>Apply here</strong></a>,
<a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fjobs.ashbyhq.com%2Ftldr.tech%2Fc227b917-a6a4-40ce-8950-d3e165357871/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/XJtOUh04jnxxpdu_cGyU70zTuIB2vX_5EBwCmKLeRUM=452" rel="noopener noreferrer" style="color: #0000EE; text-decoration: underline;" target="_blank"><strong>create your own role</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! TLDR is one of <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fwww.linkedin.com%2Ffeed%2Fupdate%2Furn:li:activity:7401699691039830016%2F/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/-1yeKqnneUCJtasePZJCene_AUVtKWMMgPWLNS_4Jgs=452" rel="noopener noreferrer" style="color: #0000EE; text-decoration: underline;" target="_blank"><strong>Inc.'s Best Bootstrapped businesses</strong></a> of 2025.
</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/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/aKd7VvZFquspNYMuEgebnuuYPxz1ogLWWlvugfjVf70=452"><span>Joel Van Veluwen</span></a>, <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fwww.linkedin.com%2Fin%2Fjennytzurueyching%2F/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/E0MI1jPmGv1x9dpkY4UjgszRNyFls6ocKRpXKEbv6gU=452"><span>Tzu-Ruey Ching</span></a> & <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fwww.linkedin.com%2Fin%2Fremi-turpaud%2F/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/7s3NoKSkTl8x4XOzBRuS0r_lkQFNHkcZDh_RzEc6pWA=452"><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/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/s965LPbNJ0HfNGxqXpl1gWkJcQ3XA9lqtYjKFNrPZu4=452">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=48058a66-736a-11f1-98dc-afc99fb97726%26pt=campaign%26pv=4%26spa=1782727223%26t=1782727672%26s=80bf988557e05fb065d53c4c49c01ebc5cdffd49ae6e8854e4d1c4a3125505fe/1/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/RoU6Pz9OLJdlT1BV0sOTFXiu3XsrJ7-QWbila0IeqEQ=452">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/CI0/0100019f12d98340-2871b539-2120-4ed3-91eb-cc57d267cc3c-000000/_UedicjTHfcE_pvprxxIlMEGwvopRp475VlprRsbKgU=452" style="display: none; width: 1px; height: 1px;">
</body></html>