Tyler Stapler

Remember that not only does shipping matter, but cleaning house. Clunky codebases with outdated code or core ideas are slow to develop in and require more bodies to just get stuff done. When you don't invest in core architecture that EVERYONE owns (Individual Codebases, Systems, Product, Management, Leadership), you break the trust and introduce delay. Yeah - solving problems means EVERYONE. I don't care what level you are or position you hold. TLDR: Core Architecture matters, but make it part of your CI/CD.

27

2 Comments

I recently presented this slide deck on "The Rise of Object Store Architectures" to ~500 engineers at Target. It's been 19 years since S3 was first released in 2006, but there has been an explosion recently in alternative data store architectures built on top of S3-compatible object stores. The slides detail why this shift is happening and what tradeoffs this architecture has for your application.

202

10 Comments

It is legitimately hard to support Google Play right now. Why? The review times are out of hand and continue to get worse. It makes Android Apps and Android itself insecure and vulnerable because Developers cannot fix things in a reasonable amount of time to get to customers, with many devs reporting well over the "7 Days" of review times. The worst part? It genuinely prevents Android from moving at Market Speed and gives iOS a real edge in the market. A week can be the difference between capitalizing on a trend or not. You cannot expect companies to support Native apps when they can't ship fixes or features to customers in a reasonable time frame or even consistent review windows. I really don't think someone at Google Play is doing the full math here on what the impact to total revenue is beyond the short-term gains in reduced staffing. When consumers have a choice and Android is perceived as consistently buggy and delayed compared to iOS, I cannot blame people for switching.

109

28 Comments

🚀 New Built-in Blue/Green Deployment for Amazon ECS: Safer, Faster, and More Reliable Releases 🚀 Containers have transformed the way development teams package and deploy applications. But deploying at scale can be tricky. Teams often end up building custom tooling just to manage deployment risks, which slows down shipping velocity. 👉 But now, there’s a better way. Starting today, Amazon ECS introduces a built-in blue/green deployment capability that helps make your application deployments safer and more consistent. With this feature, you no longer need to spend cycles building custom deployment tools. Instead, you can: ✅ Deploy faster and more confidently with automatic rollback capabilities. ✅ Ensure safer software updates by testing in production environments before exposing them to users. ✅ Easily monitor and shift traffic between blue and green environments. This feature is directly integrated into Amazon ECS—no need for additional tooling. How it works: Create a "green" environment while your existing "blue" environment handles live traffic. Test the green environment, and once ready, shift live traffic from blue to green. Amazon ECS ensures that your green environment gets validated using synthetic traffic before production traffic is routed. You can also integrate deployment lifecycle hooks (via AWS Lambda) to add custom logic for validating your deployment before moving forward. Why it matters: 🛠️ Less time spent on custom tooling. ⏱️ Faster, more reliable rollbacks when issues arise. ⚡ More frequent, seamless deployments—without the usual downtime. 📦 Pricing: This blue/green deployment capability comes at no additional cost beyond the compute resources used during deployment. 🌍 Availability: Available in all commercial AWS Regions. Ready to try it? Get started by configuring your ECS service today! 🙌 Check out the AWS documentation page for more info: https://lnkd.in/gkMAFh3A #AmazonECS #Containers #DevOps #AWS #Deployment #CloudComputing #CI_CD

20

You are going to do a System Design interview ? You want to practice ? A fun prompt that I tested with Claude 4 Sonnet and is really interesting: # System Design Interview Simulation Act as a senior system design interviewer at **[COMPANY_NAME]**, a leading **[PRODUCT_TYPE]** company. You are conducting a 45-minute system design interview with a candidate for a **[POSITION_LEVEL]** position. ## Your Role: - Present realistic system design challenges relevant to **[DOMAIN/INDUSTRY]** - Ask clarifying questions about requirements, scale, and constraints - Probe deeper into technical decisions and trade-offs - Provide constructive feedback on architectural choices - Guide the conversation like a real interviewer would - Challenge assumptions and ask follow-up questions ## Interview Structure: 1. Present 3 system design topics (I'll choose one) 2. Start with requirement gathering (5-10 minutes) 3. High-level architecture design (15-20 minutes) 4. Deep dive into specific components (10-15 minutes) 5. Discuss scalability, reliability, and trade-offs (5-10 minutes) ## Your Interviewing Style: - Ask "Why did you choose X over Y?" frequently - Request specific numbers for scale estimates - Challenge me on potential bottlenecks - Ask about failure scenarios and monitoring - Probe about data consistency and security (especially important for **[DOMAIN_SPECIFIC_CONCERNS]**) - Give real-time feedback on my approach ## Possible Topics: Please present 3 system design challenges relevant to **[COMPANY_NAME]**'s domain, such as: - **[TOPIC_1]** - **[TOPIC_2]** - **[TOPIC_3]** ## Company Context: - **Scale:** **[COMPANY_SCALE]** (e.g., "10,000+ customers with varying sizes from 10 to 50,000 employees") - **Key Challenges:** **[DOMAIN_CHALLENGES]** (e.g., "compliance, data privacy, multi-tenant architecture") - **Integrations:** **[INTEGRATION_COMPLEXITY]** (e.g., "500+ third-party apps", "multiple payment providers") ## Evaluation Criteria: Rate my performance on: - Problem decomposition and requirement gathering - System architecture and component design - Scalability and performance considerations - Data modeling and storage decisions - **[DOMAIN_SPECIFIC_CRITERIA]** (e.g., "Security and compliance awareness") - Communication and thought process ## Example Interaction Flow: Interviewer: "Design a system for [SPECIFIC_FEATURE] at [COMPANY_NAME]. We have [SCALE_CONTEXT]." Candidate: "Let me start by understanding the requirements..." Interviewer: "Good approach. What specific questions do you have about [DOMAIN_PROCESS]?" Candidate: "What are the main steps in [PROCESS]? What integrations are needed?" Interviewer: "[TYPICAL_FLOW_DESCRIPTION]. We integrate with [INTEGRATION_DETAILS]. Now, what's your approach to handle this complexity?"

21

3 Comments

I've been thinking a lot lately about being a "respectful" engineer for the web. When someone visits a site, their browser dedicates a single thread to running *our* JavaScript code and responding to *their* interactions. If our code runs too long, the browser sides with us and stops listening to them, and that's when users experience "jank" and a whole lot of frustration. We're just borrowing the main thread from the user. We need to share it, and be more respectful with it. I wrote a new post about this with some examples and practical advice: https://lnkd.in/ePJEdqcX #JavaScript #WebPerformance #Frontend

61

14 Comments

Some exciting updates are coming to POS UI extensions 🚀 This summer, we introduced new capabilities: Direct API access, Storage API and new targets on the cart line item details screen. I'm excited to preview even more exciting things coming in 2025-10 and beyond. 🎨 Polaris We're lowering the barrier to making great POS UI extensions by bringing Shopify's single UI framework, Polaris, to POS. Write highly-performant, HTML-like code that works with any framework for any app and any extension surface, including Shopify POS. Preview components and examples in the release candidate documentation: https://lnkd.in/gwciA7CT We understand this is a big change if you've previously been writing your UI code with React. Get ready with our upgrade guide: https://lnkd.in/gjQAY-pf 🗺️ Navigation Extensions can present native detail screens. Present product details, product variant details, customer details, order details, and staff details without retail staff leaving your extension and workflow. Documentation: https://lnkd.in/gvmd6etq 🛒 Cart API We're adding support for subscriptions and making tweaks to how we provide your extension with updates from the cart. Documentation: https://lnkd.in/g3FCe-kX 📚 Documentation improvements We have added a new targets overview page that provides a great overview of available targets with improved descriptions and visuals: https://lnkd.in/g6nDZmu8 🛠️ Developer Experience We're making improvements to the developer experience with better in-app extension development tools, better debugging tools, and hot reloading. Keep an eye on upcoming versions of Shopify POS as we'll be rolling out these improvements into January. 🔮 Next-gen developer platform Shopify has launched a new Dev Dashboard, and we've transitioned apps and dev stores from the Partner Dashboard. Make sure to update to the latest version of Shopify's CLI and check out the latest news: https://lnkd.in/gkGvn7Tz

84

8 Comments

Incident response is one of my favorite use-cases for demonstrating the value of building on top of the Claude Agent SDK. I had so much fun working with the talented Ben Lehrburger on a new cookbook for the SRE Incident Response Agent, which walks through building an SRE agent on the Agent SDK! A few things we cover in our cookbook: - How to give agents safe write access to infrastructure by scoping MCP tools with restricted directories, command allowlists, and validation hooks - Why clear tool descriptions drive autonomous agent behavior more effectively than elaborate prompts - How agents can synthesize across production signals like metrics, logs, alerts, and configuration to build a coherent diagnosis that no single data source could reveal on its own - How to structure human-in-the-loop workflows that separate investigation from remediation, so you control when agents act autonomously Check it out at https://lnkd.in/gdZTV85s!

235

9 Comments

🦺 🛟 Agent Safety > Model Safety LLMs didn't break your app. Agents can. As we wire LLMs to tools, browsers, payments & code the risk shifts from bad text to bad actions. Top failure modes (seen in the wild & in papers): 🚩 Prompt/Content Injection -> Tool Misuse. Malicious web pages or data instruct agents to steal secrets or make unintended API calls. This is now a primary risk for browsing and MCP - style tool agents. 🚩 Over permissioned tools & long-lived tokens. Agents get "god mode" scopes. A single injection becomes account takeover or irreversible ops. Security notes for MCP emphasise least privilege and short-lived auth. 🚩 Unsafe web autonomy. Benchmarks show web agents will attempt harmful tasks unless constrained (posting misinformation, illicit sales, etc.). You must measure misuse not just assume guardrails. 🚩 Supply-chain & retrieval poisoning. Agents trust plugins, third-party tools & indexed data that an attacker can taint. OWASP’s newest GenAI Top 10 calls this out explicitly. 🚩 Process safety gaps. NIST's GenAI Profile (AI 600-1) warns that organizations ship agents without role clarity, human-in-the-loop or incident playbooks. ✅ Model safety reduces bad text. Agent safety prevents bad state. Treat the agent like a junior SRE - with narrow permissions, audited actions and clear escalation paths. If you're building agents, what's the one control you won't ship without? Also drop a comment if you'd like me to share a follow-up post on how to tackle each of these risks. #AgenticAI #AISafety #AIAgents #GenAI #CyberSecurity #MCP #OWASP #NIST #AITrust #ResponsibleAI #AIForBusiness

5

Went to an MCP demo night yesterday and it reinforced a few gaps I've been thinking about when it comes to how we’re building with AI. 💡 There’s a lot of incredible momentum. People are vibe coding, shipping fast, and building genuinely cool MVPs. At the same time, these broader gaps stand out. 👉🏻 Gap 1: Security still isn’t part of the default build mindset I understand why security often comes later, especially when you're just trying to get something working. But when you’re building tools that are deeply integrated into data, user context, and external systems, security can’t stay optional for too long And ironically, if you're building with AI, it's easier than ever to at least set basic guardrails. 👉🏻 Gap 2: A lack of understanding around what’s actually being exposed As AI becomes more embedded into applications, things like secrets, credentials, or sensitive internal information should never be exposed in prompts or model context. That’s not always obvious, especially for people new to building AI-native apps. and this isn’t about calling anyone out. It’s just a reminder that the attack surface shifts when applications are driven by context and behavior, not just code. Overall, the energy around MCP and AI tooling is genuinely exciting. There’s a lot of thoughtful work happening, and it’s clear the ecosystem is moving fast and I’m excited to see where this goes. I just hope security becomes something we talk about earlier in the build process, not only when things scale or break. Appreciate New Enterprise Associates (NEA), Merge, Databricks, and Perplexity for hosting, and GitHub for the beautiful space. 👀 Curious how other builders are thinking about security while moving fast #mcp #founders #aitools #sf

29

Classic SRE is the only thing keeping Generative AI from being a dangerous toy. Companies are currently flooded with "AI Engineers" who can build cool prototypes but have no idea how to run them at scale, secure them, or control their costs. They are building "probabilistic software" (AI) using "deterministic ops" (traditional DevOps), and it's failing in production. Anyone agree with Gemini's take? #sre #devops #gemini

63

12 Comments