<!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 is reshaping analytics by reviving semantic layers for meaning, speed, and trust, while dashboards remain essential for core monitoring </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/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/4fZK4-6K0A9WfndZHzZ8zNEDGaRIZCHUjemL3nEBHHg=443" 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/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/2fLA1aAbssdNKZzJAUNOAsC6fEfXXAe-6GQkiapL860=443" 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=843c9e4e-0249-11f1-b72c-4770f3ede969%26pt=campaign%26t=1770289675%26s=c027a3f37c28309758940cc3e8a2166adfc377dee0bfc66c500432347638f778/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/rVIUFYJPdGKvpeAmmza6Q3b0LbDFcNPAHxtjtCKf2oY=443"><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">2026-02-05</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%2FASSxxr/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/tKzicGoanDnUU7kqc6l6dSrFtzrf9DRkdKhWlJXdZFc=443">
<span>
<strong>Lessons learned from scaling data scientists with AI (7 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Whatnot scaled its data science team by heavily leaning on LLMs and AI tools. It built custom internal tooling around strong prompting patterns, creating reusable prompt libraries, implementing strong human-in-the-loop review processes, and treating LLM outputs as strong first drafts rather than final answers, allowing a small team to support high business velocity.
</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%2Fengineeringblog.yelp.com%2F2026%2F02%2Fhow-yelp-built-a-back-testing-engine-for-safer-smarter-ad-budget-allocation.html%3Futm_source=tldrdata/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/EJ3fzQvKB-v8uP14AadJG-N33R1tbndv3m93A1ZuQow=443">
<span>
<strong>How Yelp Built a Back-Testing Engine for Safer, Smarter Ad Budget Allocation (10 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Yelp built a Back-Testing Engine to safely simulate proposed changes to its complex Ad Budget Allocation system by replaying historical campaign data through production-like logic, using tools like CatBoost ML models for outcome prediction, Scikit-Optimize for parameter tuning, and Git-submodule-integrated code to evaluate system-wide impacts without risking live advertisers.
</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%2Fhackernoon.com%2Fhow-sharechat-scaled-their-ml-feature-store-1000x-without-scaling-the-database%3Futm_source=tldrdata/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/0GXZyiWnjRQkkIfd9m80_SRtq_b4Clzg6_xQhQQnhyo=443">
<span>
<strong>How ShareChat Scaled their ML Feature Store 1000X without Scaling the Database (7 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
ShareChat engineers overhauled their ML feature store with the aim of serving 1 billion features/second on ScyllaDB. The original design failed at 1 million/sec, primarily due to inefficient data modeling and tiling. Key optimizations included serializing features with protocol buffers, refining tile aggregation intervals, adopting leveled compaction, and ultimately splitting into smaller services to increase cache hit rates (to 95%). Integrating Envoy Proxy further boosted cache hits to 98% and enabled further scaling without expanding the database cluster.
</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.youtube.com%2Fwatch%3Fv=tEIQGgS4Zus%26utm_source=tldrdata/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/YiOwjvmQhb-8Eyw-85gE31t7SefRQ2-m3VQvspJS6oQ=443">
<span>
<strong>Dashboards vs. Agents: Navigating the New Era of BI and Analytics with Mike Driscoll (54 minute podcast)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
AI is reshaping analytics by reviving semantic layers for meaning, speed, and trust, while dashboards remain essential for core monitoring even as ad hoc reporting declines. The real shift is closing the loop from insight to action, which makes strong data engineering standards and deep domain expertise more important, not less.
</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%2F32taE9/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/uNEJOWhH4OA-pYKDGF3fQN6Mgd2yDdIgFnMbiqO2qGo=443">
<span>
<strong>The Gravity of Open Standards: PostgreSQL as the Ultimate Anti-Lock-In Strategy (6 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
PostgreSQL embodies the "gravity of open standards" as the ultimate anti-lock-in strategy, thanks to its vendor-neutral community governance, strict adherence to SQL standards, open-source licensing that prevents proprietary capture, long-term backward compatibility, and design prioritizing data portability and migration freedom. This approach creates strong ecosystem "gravity" through trust, broad tooling, stable extensions, and real-world adoption.
</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.infoworld.com%2Farticle%2F4123664%2Fthree-ways-ai-will-change-engineering-practices.html%3Futm_source=tldrdata/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/U2bXV3GKb-HGBS9FKsZt-2hQpUorCuQMim5kpSiH010=443">
<span>
<strong>Three ways AI will change engineering practices (4 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
AI integration is streamlining engineering workflows by accelerating prototype development, improving architecture validation, and enhancing productivity through automated research and documentation. Teams are now required to strengthen documentation practices to maximize AI effectiveness and context accuracy, while heightened compliance and data governance are critical due to expanding privacy regulations.
</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%2Fhackernoon.com%2Fenterprises-dont-have-an-ai-problem-they-have-an-architecture-problem%3Futm_source=tldrdata/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/Z6pTKFvQfGOh80h5BwlhMIWAKQEYHMCTvlYBtLZwOKs=443">
<span>
<strong>Enterprises Don't Have an AI Problem. They Have an Architecture Problem (4 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Many organizations mistake deploying chatbots, RAG systems, or workflow automation for true enterprise AI, but these are isolated tools, not an enterprise capability layer. Effective enterprise AI demands architected integration with clear business outcomes, robust governance, consistent data ownership, and alignment with frameworks like TOGAF. Without architectural discipline, organizations accumulate technical debt, risking fragile, unscalable systems instead of realizing AI's transformational potential.
</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%2Fhydrolix.io%2Fsolutions%2Fbot-insights%2F%3Futm_source=Newsletter%26utm_medium=Email%26utm_campaign=TLDR/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/3pxPBcNwY86Nu9o5p4qQOXK8hEpI_U_N1QUv6XgVXIQ=443">
<span>
<strong>The $5 Million Bots Bill (Sponsor)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Most web traffic is driven by bots, and it's crushing companies' budgets. (One client found Hydrolix after bot traffic bypassed their firewall, hit origin servers, and triggered >$5million overcharges.) Hydrolix accurately classifies human and bot traffic in real time, identifying good bots, AI scrapers, impersonators, emerging threats, etc - then mitigates them instantly. See how it works.
</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%2Fegeominotti%2Fbunqueue%3Futm_source=tldrdata/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/jfRD92PBaI7ENzQ9vmtR4LiH-H5oMxYSK45vhEt5-hk=443">
<span>
<strong>bunqueue (GitHub Repo)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
bunqueue is a Bun-first job queue built on SQLite (WAL) that aims for very low latency and simple ops: no Redis, no external services, and a BullMQ-compatible API. It supports embedded (in-process) or TCP server mode, and includes production features like retries/backoff, stall detection, DLQs, rate limiting, cron, metrics, and S3 backups.
</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%2Fsqldef%2Fsqldef%3Futm_source=tldrdata/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/KZN_KkFRH7m_vARB0vCNu2-boyl_5tjSl_LEev3srt8=443">
<span>
<strong>sqldef (GitHub Repo)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
sqldef is a single-binary, idempotent schema management tool that lets you define your entire database schema in plain SQL and safely apply changes by diffing desired vs current state. It supports MySQL, PostgreSQL, SQLite, and SQL Server, works online or offline, generates deterministic DDL, and fits cleanly into CI/CD without custom DSLs or migration files.
</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.cncf.io%2Fblog%2F2026%2F02%2F02%2Fopentelemetry-collector-vs-agent-how-to-choose-the-right-telemetry-approach%2F%3Futm_source=tldrdata/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/sFGAQ6Zig4JIHAA-z1JrEwO3cvXsLxo7Vmy6y-9Giqc=443">
<span>
<strong>OpenTelemetry Collector vs agent: How to choose the right telemetry approach (3 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
OpenTelemetry's rapid adoption is redefining observability for cloud-native architectures, with its Collector serving as a central hub for aggregating, transforming, and exporting telemetry, and its agent acting as a lightweight sidecar for local data capture. Centralized control and scalability depend on the Collector, while low-overhead, source-proximate telemetry favors the agent.
</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%2Fblog.bytebytego.com%2Fp%2Fhow-grab-built-a-vision-llm-to-scan%3Futm_source=tldrdata/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/Gw1j28p3RJVFo5WnnAbIZMVnNKjYFwLSO9QcB642PsE=443">
<span>
<strong>How Grab Built a Vision LLM to Scan Images (10 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Grab built a custom 1B-parameter Vision LLM from scratch to improve customer ID verification by accurately extracting information from diverse, real-world document images in Southeast Asia, overcoming the shortcomings of general-purpose vision models through a specialized four-stage training pipeline (projector alignment, vision enhancement, language-specific training, and task-focused fine-tuning) using synthetic/real data generated by an auto-labelling framework.
</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%2Fhackernoon.com%2Fhow-ai-agents-helped-migrate-a-data-lake-from-snowflake-to-bigquery%3Futm_source=tldrdata/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/tvyFyN-ogDjM4-PbPgjxPhLxDdC4pxr24UjZdUIgdnA=443">
<span>
<strong>How AI Agents Helped Migrate a Data Lake From Snowflake to BigQuery (4 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Migrating a data lake between data platforms requires comprehensive pipeline refactoring and translation of complex, layered SQL logic. Leveraging generative AI agents and centralizing refactoring guidelines and workflow in a structured AGENTS.md enables efficient rewriting and validation of hundreds of scripts and database objects where traditional NLP and other string-matching methods fail.
</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%2Flinks.tldrnewsletter.com%2FPXIjyq/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/pbA7cBDiSaSHmzmHMt6KKuEAIznA978aJE64dD3g8-E=443">
<span>
<strong>High Performance Map Transformations in PySpark (4 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
As semi-structured JSON becomes more common, PySpark's native map functions enable fast, declarative transformations at scale without costly explode or UDF patterns.
</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%2Fragaboutit.com%2Fthe-small-language-model-awakening-why-enterprise-rag-is-abandoning-foundation-models-for-domain-specific-precision%2F%3Futm_source=tldrdata/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/h5_2Llub9n4i_ijDlbS-vg8HJw5vVBlsNeqSNg-ynzk=443">
<span>
<strong>The Small Language Model Awakening: Why Enterprise RAG Is Abandoning Foundation Models for Domain-Specific Precision (8 minute read)</strong>
</span>
</a>
<br>
<br>
<span style="font-family: "Helvetica Neue", Helvetica, Arial, Verdana, sans-serif;">
Enterprises are moving from general-purpose LLMs to domain-specific SLMs, which deliver higher accuracy, stronger governance, and far lower inference costs in regulated industries like life sciences and banking.
</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/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/9CAa2Dx7Avy72OEpNewjZpa32wxqnZOO4ArfS4LEBfw=443"><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/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/zk1iNPsBC2nkdGW1LcLyrd7uPBRJa_R8RYcAfBHcw3Q=443" 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-d3e165357871/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/WGXpxdDmJDT9pRGm2Jk-jg8nsqoE6HZNIeRGE4dQa_U=443" 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/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/JmOb2YcvB3cwGtGaISx6kDfoZaCGxTzkrCtluPycuSc=443" 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/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/u4xREor-wfhVNeGZTLIprDj830kDl7PvO3rXe0VEX6c=443"><span>Joel Van Veluwen</span></a>, <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fwww.linkedin.com%2Fin%2Fjennytzurueyching%2F/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/VHCwhq-GcRJFy59_QPP2fsgBD79V0rK_9ozv6zaxn6k=443"><span>Tzu-Ruey Ching</span></a> & <a href="https://tracking.tldrnewsletter.com/CL0/https:%2F%2Fwww.linkedin.com%2Fin%2Fremi-turpaud%2F/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/K7DS9vKoXbVY4rmksbynTeJZgDcngopYvMzSNvdOdOM=443"><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/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/e7b0lhclv-l-w4K31X3bi7XlhvxR-nO7yffwQY3o-D8=443">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=843c9e4e-0249-11f1-b72c-4770f3ede969%26pt=campaign%26pv=4%26spa=1770289288%26t=1770289675%26s=b03e1913bcbccb2ebff607ef8102ebc8bcae6967fd0737707a9e4b790e6f3fa0/1/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/SLgBvhE9WRRFWiPpEL1b-fgoDnDzqcx0r2qc_aLmVgA=443">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/CI0/0100019c2d7cbe4e-89b851c9-9dbb-4657-96b3-7ce2c6f97481-000000/1Wfy3WKOilk-YbIoSF8yWLlClcVDhwii8M07--Tk_JE=443" style="display: none; width: 1px; height: 1px;">
</body></html>