Articles by Jason
-
pg_sage - an agentic postgres dba
pg_sage - an agentic postgres dba
I have been working on databases since 1998. When LLMs first launched, they were horrible with SQL.
4 Comments
Activity
4K followers
-
Jason Massie reposted thisOn today’s Q1 2026 earnings call, Sundar Pichai recapped a terrific start to the year, with our AI investments and full-stack approach lighting up every part of the business. Check out all the highlights ⬇️Jason Massie reposted thisQ1 earnings are in: 2026 is off to a terrific start thanks to our partners + employees around the world. Our AI investments and full stack approach are lighting up every part of the business: Search queries are at an all-time high with AI continuing to drive usage. Google Cloud revenue grew 63%, Gemini models have incredible momentum, and it was our strongest quarter ever for consumer AI subs, driven by the Gemini app. Overall the number of paid subscriptions has now reached 350 million, with YouTube and Google One being the key drivers. Beyond the topline numbers - a few other highlights that show the progress: - Faster results: Even as we’ve brought new AI features into our results page, we’ve reduced Search latency by more than 35% over the past 5 years. - Enterprises’s agentic era is here: Over the past 12 months, 330 Google Cloud customers each processed over 1T tokens, while 35 reached the 10T token milestone. And in Q1, Gemini Enterprise paid monthly active users grew 40% quarter-over-quarter. - Twice the rides: Waymo has doubled the amount of fully autonomous rides in less than a year, now operating in 11 US cities. Much more to come at I/O, 20 days away! https://lnkd.in/gzRgFqAB
-
Jason Massie shared thispg_sage v1 is up on github. https://lnkd.in/g85Ei5cs Over the past few days, pg_sage moved from a Postgres observability companion toward a real autonomous DBA control plane. Who needs another dashboard? The big shift is that database work now flows through Cases: a single operational queue that connects findings, incidents, migration risks, proposed actions, approvals, execution history, rollback guidance, and verification evidence. The goal is not “chat with your database”; it is a guarded, auditable system where deterministic checks and typed action contracts let an LLM reason inside clear boundaries, while DBAs keep visibility into what happened, why it matters, what would be automated, and where human approval is still required. New features: Unified Cases work queue for findings, incidents, forecasts, query hints, migration risks, and action history Shadow Mode reporting to show avoided toil before enabling autonomous execution Provider readiness matrix for Cloud SQL, AlloyDB, RDS, Aurora, and self-managed Postgres DDL safety preflight with PR/CI-ready migration output, rollback or forward-fix guidance, and verification SQL Incident playbooks for lock blockers, runaway queries, connection exhaustion, WAL/replication risk, standby conflicts, and sequence exhaustion Vacuum, bloat, and freeze autopilot with guarded maintenance candidates and IO-saturation gates Query tuning beyond hints, including rewrites, broken-hint retirement, CREATE STATISTICS, parameterization plans, and reviewed role-level work_mem changes Trust-ramped action execution with approval gates, cooldowns, action expiration, rollback metadata, and verification state Fleet-safe behavior across multi-database settings, detail views, mutations, and action execution Hardened admin, notification, settings, fleet, and legacy-route workflows with browser/API/database verification On side note: This version switched from Claude to Codex. I really disliked OpenAI. However, they are currently better. There is truly no moat and switching platforms is super easy.GitHub - jasonmassie01/pg_sage: An Agentic PostgreSQL DBA— monitors, analyzes, and optimizes any PostgreSQL 14-18 database with LLM-powered actionsGitHub - jasonmassie01/pg_sage: An Agentic PostgreSQL DBA— monitors, analyzes, and optimizes any PostgreSQL 14-18 database with LLM-powered actions
-
Jason Massie reposted thisJason Massie reposted thisFrom its inception, 𝐁𝐢𝐠𝐭𝐚𝐛𝐥𝐞 has been the engine behind applications that redefined expectations for scale, reliability, and performance—from Google Search and Maps to the largest Google Cloud customers. Last week at Google Cloud Next, we saw this evolution continue: • 𝐆𝐞𝐦𝐢𝐧𝐢: Sundar Pichai shared how Gemini leverages Bigtable to power the next generation of AI. • 𝐀𝐧𝐭𝐡𝐫𝐨𝐩𝐢𝐜: The team joined us to discuss how they manage 100s of petabytes of data on Bigtable to support Claude. With the introduction of a new in-memory tier and expanded query capabilities, 𝐁𝐢𝐠𝐭𝐚𝐛𝐥𝐞 remains at the forefront of the AI revolution. If you’re building your next big AI project, we have the foundation you need. https://lnkd.in/gn4i-YtTScaling intelligence: How Anthropic secures Claude with BigtableScaling intelligence: How Anthropic secures Claude with Bigtable
-
Jason Massie reposted thisJason Massie reposted thisThe reactive era of the System of Intelligence is over. Enterprises are looking for agents to be proactive, understand their business, and be able to take action on their behalf. Today we are unveiling the Agentic Data Cloud - the era of the System of Action has begun. You need a Cloud that can support going from human scale to agent scale in a way that is cost efficient and effective. You need a system that is no longer reactive but proactive, yet trustworthy. You need a system that understands your business semantics and transitions from data to knowledge. Welcome to the Agentic Data Cloud! — built on Google’s differentiated AI infrastructure, models, and data systems. https://lnkd.in/gwseMvDrArchitecting the Agentic Data Cloud | Google Cloud BlogArchitecting the Agentic Data Cloud | Google Cloud Blog
-
Jason Massie reposted thisJason Massie reposted thisI am excited to host #Anthropic's David Soria Parra, co-creator of MCP, for a Fireside Chat at Google Cloud Next! 🚀 We're diving deep into the Model Context Protocol (MCP) and the future of Agentic AI. You won't want to miss this session—details below: Session: Power intelligent agents with AI-native databases Fireside Chat Topic: Bridging the gap between LLMs and real-world data ☁️ April 23 | 🕒 1:15 PM 📍 Mandalay Bay F, Session BRK2-021 Register here: [https://lnkd.in/gpNFa-RU] If you’re attending Next, add this to your schedule! I can't wait to see you there. 🤝 #Anthropic #GoogleCloud #GenerativeAI #MCP #AIagents
-
Jason Massie reposted thisJason Massie reposted thisDo you vibe code? Are you attending Google Cloud Next ‘26? Join me for a 1:1 coffee chat to discuss vibe coding and databases! DM me and can figure out when/where. Thanks! #GoogleCloudNext #vibecoding #postgresql #sql
-
Jason Massie reposted thisJason Massie reposted thisAre you building for the agentic era? 🤖 To deliver truly relevant, accurate AI experiences at scale, you need a unified, real-time data foundation. I'm incredibly excited for our upcoming #GoogleCloudNext session on April 22nd: "What's new in Google Cloud databases for the agentic era." We will showcase the latest features designed to remove friction from the developer experience, simplify database management, and help you seamlessly blend multimodal AI data with your real-time operational workloads. Check out the session catalog and add it to your schedule! ➡️ goo.gle/4dUPBCy #GoogleCloud #Databases #Developers #AI Raj Pai Samuel Moses Wenzhe Cao
-
Jason Massie reposted thisJason Massie reposted thisReady to see what AI search with autonomous agents looks like in practice? 🚀 I’m thrilled to be heading to Las Vegas for Google Cloud Next ‘26. Alan Li from Google, Sebastián Queirolo from Mercado Libre and I are presenting a breakout session on the topic "AlloyDB AI in action: powering next-gen search and autonomous agents” on Thursday, April 23, at 4 pm. We will dive into industry-leading vector search with AlloyDB, and unveil our latest advancements across various pillars: AI powered search at scale, hybrid search, Gemini-based AI functions, and natural language to SQL (our new QueryData tool) to build agentic applications. The era of "Agentic AI" isn't just coming - it’s already here on AlloyDB! If you're attending Google Cloud Next, I’d love to see you there! 👇 Save your spot (link to the session in comments). #GoogleCloudNext #AlloyDB #GenerativeAI #Databases #AIagents #AlloyDBAI #PostgreSQL
-
Jason Massie reposted thisJason Massie reposted thisI’m excited to share that I'll be speaking at Google Cloud Next 26 on April 22nd, 2026! Join me for "NoSQL for modern apps and AI: The future of Memorystore, Firestore, and Bigtable". We'll be unveiling exciting innovations in NoSQL databases launching at Google NEXT. Plus, you'll hear from Chethana S. as she shares how Palo Alto Network leverages Google's NoSQL databases for their mission-critical applications. Register now: http://shortn/_LgXHdSlUgB. #GoogleCloudNext #NoSQL #AI #Firestore #Bigtable #Memorystore #GoogleDatabases
-
Jason Massie liked thisJason Massie liked thisTo my network across the Twin Cities and Minnesota: Join us in #Minneapolis on June 9th for Agentic Work: Live + Labs, an exclusive half-day event hosted by Google. We’re unpacking how Gemini Enterprise and Google Workspace are shifting from simple prompts to semi-autonomous agents that drive real business value. Register: https://google.smh.re/5UPO What to expect: 🚀 Keynotes & Strategy: How to deploy secure, governed agent collections. 🛠 Workshop: Get hands-on with our tools and learn AI-first leadership strategies. 🤝 Real-world Insight: Customer panels and networking with peers driving the next wave of AI. 📍 Minneapolis: Tuesday, June 9 | 12:00 PM – 5:00 PM #Google #AgenticAI #GoogleWorkspace #Gemini #FutureOfWork #MinneapolisTech #GenerativeAI
-
Jason Massie liked thisJason Massie liked thisHow Sundar Pichai Pushed Google To the Front of the AI Race ". ..While analysts called for his resignation, Pichai remained calm; he had been lying in wait for this moment for a decade. In 2016, he had declared Google would be an “AI-first company,” and began cultivating a series of projects—custom chips, Cloud, YouTube, and deep AI research—that seemed to have nothing to do with Google’s core search product. ... All of these bets have paid off, and then some. Google DeepMind, the company’s AI-research lab—led by Nobel Prize–winning CEO Demis Hassabis—forged several key breakthroughs that catapulted Google’s Gemini model to the top of many capability leaderboards. Gemini now accounts for a quarter of AI traffic worldwide, up from 6% a year ago, according to Similarweb. Google has quietly introduced millions of people to AI through everyday products: search, image-generation tools like Nano Banana, video editing on YouTube, research assistance via NotebookLM, translation through Google Translate, and autonomous driving with Waymo. At the same time, its Cloud division has boomed, powering a wave of businesses entering the AI economy. "TIME100 Most Influential Companies 2026: How Sundar Pichai Pushed Google To the Front of the AI RaceTIME100 Most Influential Companies 2026: How Sundar Pichai Pushed Google To the Front of the AI Race
-
Jason Massie liked thisJason Massie liked thisThe agentic DBA is here. Amazing stuff here at Google Cloud Next and learning about the AI native database platform. Some much to learn but really fascinating.
-
Jason Massie liked thisJason Massie liked thisRecognized as a Google Cloud subject matter expert for my contributions in creating technical content in 2026. #googlecloudlearningTechnical Content Subject Matter Expert Badge was issued by Google Cloud to Celia Antonio.Technical Content Subject Matter Expert Badge was issued by Google Cloud to Celia Antonio.
-
Jason Massie liked thisJason Massie liked this🚀 What an incredible time on stage at Google Cloud Next 2026! We just wrapped up Cloud Next 2026 and shared a massive wave of updates for Memorystore. If you're building real-time, context-aware AI applications, you already know that disk-level latency simply won’t cut it, you need an in-memory tier that delivers data at lightning speed. To show what this looks like at true internet scale, we were incredibly thrilled to have the engineering team from Snap Inc. join us. The absolute highlight? 🔥 Snap achieved 30% more throughput (QPS) per node using Memorystore for Valkey compared to their self-managed Kubernetes deployment: running the exact same workload on the exact same machine shape! Memorystore for Valkey now proudly powers some of the most critical workloads at Snap. That is game-changing performance, but it was just one part of what we presented. We unveiled a whole suite of innovations to help you supercharge your data infrastructure: ⚡ Memorystore for Valkey 9.0: We are delivering next-gen speed, including up to 40% more throughput for pipelined workloads and a staggering 300% faster TLS connection acceptance. 🏗️ Massive Scale Expansion: We introduced six brand-new node sizes, scaling all the way up to 110 GB per node for massive AI workloads. 🔒 Enterprise Trust & Flexibility: We rolled out Flexible Certificate Authorities, Token Authentication, and seamless online migrations from self-managed setups. Granular ACLs will be launching soon! Still self managing or on a legacy service? Every hour your engineering team spends patching, tuning, and scaling self-managed clusters is an hour lost on training and optimizing your AI models. Stop managing infrastructure. Start focusing on innovation. Let’s build something incredible. Join the top 95% of Google Cloud customers and start using Memorystore today. #GoogleCloud #CloudNext2026 #Valkey #GenerativeAI #DataEngineering #MachineLearning #TechUpdates #Memorystore
Experience
-
open source
-
-
-
-
-
-
-
-
-
-
Licenses & Certifications
-
Professional Cloud Database Engineer
Google Cloud
Recommendations received
4 people have recommended Jason
Join now to viewOther similar profiles
Explore more posts
-
Arun Seetharaman
This is my favorite kind of improvement: small surface area, low blast radius, and disproportionately high architectural return. It’s not flashy, but it’s how most real-world systems actually move forward. You get better resilience and scalability without ripping through business logic or reworking data access patterns. The connection layer becomes aware of topology, roles, and authentication, exactly where that complexity should live.
-
Robert Armstrong
Did you know that the AWS command line tools use US-East-1 as their default region? If you don't specify a region or data center, your server, lambda, app, etc. are all running in US-East-1. Just thought I would remind folks to set their region in their command line and scripts.
1 Comment -
ReleaseTEAM, Inc.
Keeping your #software supply chain safe starts with what you bring in. JFrog Curation helps teams automatically vet open-source packages before they enter your environment—catching malicious, vulnerable, or non-compliant components early. The result? Cleaner builds, fewer surprises, and more confidence in every release. #OpenSource
-
Feng He
Stop teaching CodeDeploy for ECS Blue/Green! 🛑 AWS ECS has fundamentally changed deployment. It now natively supports Blue/Green and Canary/Linear strategies, eliminating the dependency on AWS CodeDeploy and drastically simplifying your CI/CD pipelines. But how do you train your team on this new, simplified flow? My open-source project, https://lnkd.in/gXeiW3iy , provides clear, animated visual guides. This project was built using #AWSKiro to bring these complex flows to life! Use it to: ✅ Instantly explain the native ECS traffic shifting process. ✅ Accelerate onboarding and unify team knowledge. ✅ Future-proof your team's skills with the latest, simplest AWS practices. Stop wading through documentation. Start seeing the simplified flow! 👉 View the animations and simplify your team's training: https://lnkd.in/gEeJ_iND Hit the Star ⭐ if your team is excited about simpler cloud deployments! #AWSECS #DevOps #CloudNative #CI_CD #AWSKiro #OpenSource
2 Comments -
Joanna Podgoetsky
If you’re using the SQL Analytics Endpoint over Lakehouse data, file health and maintenance really matters. Christopher Finlan's latest blog does a great job walking through how to ensure your data is in the right shape for fast SQL querying. Check it out https://lnkd.in/gqt4CE4x #MicrosoftFabric #Data #Analytics #SQL #DataEngineering
1 Comment -
Suhas Kajbaje
Not all managed databases are created equal. 🎯Let me break down what actually matters when you're choosing between RDS MySQL and Aurora MySQL. The Architecture That Changes Everything RDS MySQL? Traditional storage architecture. Your data lives on EBS volumes attached to your instance. Aurora MySQL? Completely different game. It separates compute from storage, using a shared, distributed storage layer that automatically scales. Your data gets written to 6 copies across 3 Availability Zones simultaneously. That's not just redundancy—that's peace of mind. What This Means for Your Applications: ✅ Recovery Point Objective (RPO) under 1 second ✅ Recovery Time Objective (RTO) under 30 seconds ✅ Read replicas share the same storage (no lag from replication) ✅ Up to 15 read replicas vs RDS's 5 replicas limit ✅ 5x throughput of standard MySQL ✅ Automatic storage scaling up to 128TB Here's what surprised me most: Aurora can lose 2 copies of data without affecting write availability and 3 copies without affecting read availability. Try doing that with RDS. Aurora Global Database vs Regular Aurora ✅ Regular Aurora = perfect for single-region applications where you need high availability and performance. ✅ Aurora Global Database = your disaster recovery superhero. It replicates data across regions with dedicated infrastructure, typically with less than 1 second of lag. Secondary regions can be promoted in under 1 minute. When do I use which? Use Regular Aurora when: ✅ Single region deployment is sufficient ✅ You need HA within a region ✅ Budget is a primary concern Use Aurora Global Database when: ✅ You have a global user base ✅ Disaster recovery is critical ✅ You need read performance close to users worldwide ✅ Compliance requires multi-region data residency The catch? Cost. Aurora costs more than RDS, and Global Database adds another layer. But when you factor in the operational overhead you're avoiding and the performance gains you're getting, it often makes sense. What's your experience with Aurora? Have you made the switch? Drop your thoughts below. 👇 Follow me on LinkedIn: https://lnkd.in/dkeKb6cc #AWS #Aurora #MySQL #DatabaseArchitecture #CloudComputing #DevOps #CloudArchitecture #RDS #DisasterRecovery #ThePracticalArchitect
2 Comments
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top contentOthers named Jason Massie
47 others named Jason Massie are on LinkedIn
See others named Jason Massie