<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Alex Z. Li on Medium]]></title>
        <description><![CDATA[Stories by Alex Z. Li on Medium]]></description>
        <link>https://medium.com/@Galaxy_Watcher?source=rss-47c0a821df07------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*2LMjzMhp3uZU2vCzOyL9Xw.jpeg</url>
            <title>Stories by Alex Z. Li on Medium</title>
            <link>https://medium.com/@Galaxy_Watcher?source=rss-47c0a821df07------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Wed, 18 Nov 2026 06:07:06 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@Galaxy_Watcher/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[How I Cut My AWS Bedrock Bill by 63%]]></title>
            <link>https://towardsaws.com/how-i-cut-my-aws-bedrock-bill-by-63-5a55bc43597b?source=rss-47c0a821df07------2</link>
            <guid isPermaLink="false">https://medium.com/p/5a55bc43597b</guid>
            <category><![CDATA[cost-optimization]]></category>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[bedrock]]></category>
            <dc:creator><![CDATA[Alex Z. Li]]></dc:creator>
            <pubDate>Tue, 17 Feb 2026 00:26:25 GMT</pubDate>
            <atom:updated>2026-02-18T11:12:38.618Z</atom:updated>
            <content:encoded><![CDATA[<h3>Context</h3><p>Running AI workloads on AWS Bedrock can get expensive fast. I operate a multi-channel AI assistant system using Claude models — it handles daily automation, tasks tracking, content generation, and analysis across Slack channels.</p><p>February bill was trending towards ~<strong>$3.9k/month</strong> running Claude Sonnet 4.5 for everything — including simple tasks that don’t need frontier-level reasoning.</p><h3>Why Optimise?</h3><p>Most Bedrock users default to a single model for all requests. The problem:</p><ul><li><strong>85% of workloads are simple</strong> — heartbeats, cron jobs, weather checks, basic Q&amp;A</li><li><strong>10% are moderately complex</strong> — code generation, document analysis</li><li><strong>5% are genuinely hard</strong> — deep reasoning, strategic analysis</li></ul><p>Using Sonnet ($3.00/M input tokens) for a weather check is like taking a Ferrari to the supermarket.</p><h3>The 85/10/5 Strategy</h3><p>Match model capability to task complexity:</p><ul><li><strong>Claude Haiku 4.5</strong> ($1.10 / $5.50 per M) — Daily tasks, cron jobs, Q&amp;A — 85% of traffic</li><li><strong>Claude Sonnet 4.5</strong> ($3.00 / $15.00 per M) — Complex analysis, coding — 10% of traffic</li><li><strong>Claude Opus 4.6</strong> ($3.00 / $15.00 per M) — Deep reasoning, strategic work — 5% of traffic</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_LISum4G3WlSpNTIPC1XZg.png" /></figure><p><strong>Implementation:</strong> Set Haiku4.5 as default, keep Sonnet and Opus as manual overrides. No smart routing — a cheap default with manual switch is simpler and more predictable than automated model selection.</p><h3>Key Learnings</h3><h3>1. Use Cost Explorer, Not CloudWatch</h3><p>CloudWatch metrics are per-model. Switch models and your dashboard goes blind.</p><p>Cost Explorer with SERVICE = &#39;Amazon Bedrock Service&#39; gives you the billing truth across all models. Group by USAGE_TYPE to see per-model breakdown as example:</p><pre>USE1-Claude4.5Haiku-cache-read-input-token-count<br>USE1-Claude4.5Sonnet-cache-read-input-token-count<br>USE1-Claude4.6Opus-cache-read-input-token-count</pre><h3>2. Cache Matters — But Has Limits</h3><p>Bedrock prompt caching saves ~90% on cached tokens.</p><p>Key constraints:</p><ul><li><strong>TTL is ~5 minutes</strong> — requests further apart get cache misses</li><li><strong>Cache is per-model</strong> — switching models invalidates all cached context</li><li><strong>Dynamic system prompts kill cache</strong> — timestamps or random content in prompts means constant misses</li></ul><p>Stabilising on one default model (Haiku) improved my cache hit rate from 60% to 85%+.</p><h3>3. Build Visibility Early</h3><p>I built a daily cost dashboard that breaks down spend by model (stacked bar chart).</p><p>This immediately caught issues — like sub-agents still running on Sonnet after I’d switched the default to Haiku.</p><p><strong>Don’t hardcode model IDs in monitoring scripts.</strong> Query all models or use Cost Explorer. You will switch models; your tools need to survive that.</p><h3>Results</h3><ul><li><strong>Sonnet-only</strong> (week1 ): avg <strong>$89.90/day</strong></li><li><strong>Haiku-default</strong> (week 2): avg <strong>$49.09/day</strong> (↓ 45%)</li><li><strong>Low-activity day</strong>: <strong>$26.36</strong> (↓ 71%)</li></ul><p>Monthly projection: <strong>$2,697 → $1,473</strong> — saving <strong>$14,688/year</strong>.</p><p>Quality impact: For 85% of tasks, no noticeable difference. Haiku 4.5 is remarkably capable. The gaps show up in multi-step reasoning and nuanced analysis — for those, switch to Opus manually.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*bUNbGcLApfjG2EZL3hI1Ow.png" /></figure><h3>What’s Next</h3><p>The Bedrock model landscape is evolving fast — GLM 4.7 Flash at $0.20/M, Kimi K2.5 at $0.60/M.</p><p>The next optimisation wave may be mixing model providers and closely evaluate the capabilities.</p><h3>Wrap Up</h3><p>The playbook is straightforward:</p><ol><li><strong>Measure first</strong> — Cost Explorer with USAGE_TYPE grouping</li><li><strong>Default cheap</strong> — Haiku handles 85% of workloads</li><li><strong>Manual override</strong> — switch up for complex tasks only</li><li><strong>Stabilise cache</strong> — one default model, no dynamic prompts</li><li><strong>Build visibility</strong> — daily dashboard with model breakdown</li></ol><p>Don’t pay frontier prices for commodity tasks.</p><h3>Resources</h3><ul><li>AWS Bedrock Pricing — <a href="https://aws.amazon.com/bedrock/pricing/">https://aws.amazon.com/bedrock/pricing/</a></li><li>Cost Explorer API — <a href="https://docs.aws.amazon.com/cost-management/latest/userguide/ce-exploring-data.html">https://docs.aws.amazon.com/cost-management/latest/userguide/ce-exploring-data.html</a></li><li>Bedrock Prompt Caching — <a href="https://docs.aws.amazon.com/bedrock/latest/userguide/prompt-caching.html">https://docs.aws.amazon.com/bedrock/latest/userguide/prompt-caching.html</a></li></ul><h3>About Writer ✍️</h3><p>Alex Li is a Cloud Architect at AWS who writes about AWS Cloud and AI infrastructure.</p><p>LinkedIn: <a href="https://www.linkedin.com/in/alexzige/">https://www.linkedin.com/in/alexzige/</a></p><p>Happy reading!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5a55bc43597b" width="1" height="1" alt=""><hr><p><a href="https://towardsaws.com/how-i-cut-my-aws-bedrock-bill-by-63-5a55bc43597b">How I Cut My AWS Bedrock Bill by 63%</a> was originally published in <a href="https://towardsaws.com/">Towards AWS</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Simplifying Kubernetes Management with Amazon EKS Auto Mode]]></title>
            <link>https://towardsaws.com/simplifying-kubernetes-management-with-amazon-eks-auto-mode-21bb27b34bd7?source=rss-47c0a821df07------2</link>
            <guid isPermaLink="false">https://medium.com/p/21bb27b34bd7</guid>
            <category><![CDATA[kubernetes]]></category>
            <category><![CDATA[aws-lambda]]></category>
            <category><![CDATA[k8s]]></category>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[amazon-web-services]]></category>
            <dc:creator><![CDATA[Alex Z. Li]]></dc:creator>
            <pubDate>Mon, 10 Feb 2025 02:58:40 GMT</pubDate>
            <atom:updated>2025-02-11T11:25:55.237Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*7j-O9xp7A-K_GADLgrP59A.png" /></figure><h3>Context</h3><p>Managing Kubernetes clusters manually is complex. Teams must handle infrastructure provisioning, scaling, security patching, and version upgrades — tasks that demand deep expertise and divert focus from core applications.</p><p><strong>Amazon EKS Auto Mode <em>(</em></strong><em>released in Re-invent 2024)</em>, addresses this by automating Kubernetes operations, enabling teams to deploy and scale containerised workloads without managing the underlying infrastructure.</p><h3>Why EKS Auto Mode?</h3><p>Traditional Kubernetes management requires ongoing effort:</p><ul><li><strong>Operational Overhead</strong>: Manually scaling nodes, updating add-ons, and configuring networking.</li><li><strong>Security Risks</strong>: Ensuring nodes comply with security best practices (e.g., patching, immutable AMIs).</li><li><strong>Cost Inefficiency</strong>: Over-provisioning resources or failing to terminate idle instances.</li></ul><p>EKS Auto Mode eliminates these by delegating cluster operations to AWS, letting teams prioritise application development and its life-cycle.</p><h3>What’s EKS Auto Mode?</h3><p>EKS Auto Mode is an <strong>AWS-managed Kubernetes service</strong> that automates both the <em>control plane</em> and <em>data plane </em>operations. Unlike standard EKS, it removes the need to configure nodes, standard AWS add-ons, or networking. Key features include:</p><ul><li><strong>Managed Infrastructure</strong>: AWS provisions EC2 instances, VPC CNI, coreDNS, load balancers, and storage.</li><li><strong>Immutable Nodes</strong>: Nodes use AWS-curated AMIs with automated rotation (replaced up to every 21 days).</li><li><strong>Integrated Tooling</strong>: Pre-installed add-ons (CoreDNS, VPC CNI) run as system processes, not pods, for stability.</li></ul><h3><strong>EKS Architecture &amp; Shared responsibility mode</strong></h3><p>Amazon EKS Auto Mode operates under a <strong>shared responsibility model</strong>, where AWS manages the underlying infrastructure and core Kubernetes components, while customers retain responsibility for securing their applications and workloads. AWS handles control plane availability, automated node provisioning/scaling, security patching of OS/Kubernetes components, and enforcing immutable node configurations (e.g., read-only filesystems, automated node rotation). Customers, however, must secure containerized applications, define IAM roles and pod security policies, encrypt sensitive data, ensure compliance, and monitor workload performance. This division allows Customer to focus on application logic while AWS ensures the cluster’s foundational reliability and security.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0lpvJK7-17x0kYtgxwws2Q.png" /><figcaption>EKS Architecture Previously</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*gOSL8Mb1EMUGWyNQU2cCsA.png" /><figcaption>EKS Architecture with Auto Mode</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YxjECQCCE0aEZ7aJqmj6Iw.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*SptiRE1SnaSOkaJYHkQVXg.png" /></figure><h3>Key Benefits with EKS auto mode</h3><p><strong>Less infrastructure Management</strong></p><ul><li>AWS handles node provisioning, scaling (via Karpenter), and termination.</li><li>Auto-heals unhealthy nodes and distributes workloads across Availability Zones.</li></ul><p><strong>Cost Optimisation</strong></p><ul><li>Dynamically scales resources to match demand.</li><li>Terminates under-utilised nodes and selects cost-efficient EC2 instances.</li></ul><p><strong>Enhanced Security</strong></p><ul><li><strong>Immutable AMIs</strong>: Nodes use locked-down, AWS-patched images.</li><li><strong>Automated Rotation</strong>: Nodes are replaced every 21 days to enforce compliance.</li><li><strong>No SSH Access</strong>: Reduces attack surfaces by blocking direct node access.</li></ul><p><strong>Simplified Networking</strong></p><ul><li>Auto-configures VPC CNI for pod networking and provisions Elastic Load Balancers (ELBs) for services.</li></ul><h3>Key Considerations</h3><ul><li><strong>Additional Cost:</strong> EKS auto mode managed instance has a slightly high costs applied on top of instance types.</li><li><strong>No Custom AMIs</strong>: Nodes use AWS-managed images; custom OS/software requires standard EKS. Auto Mode with AWS managed bottlerocket AMI.</li><li><strong>Migration considerations: </strong>Combining EKS Auto Mode with existing node groups (e.g., Karpenter) requires careful planning.</li></ul><h3>Wrap Up</h3><p>Amazon EKS Auto Mode redefines Kubernetes management by automating infrastructure, scaling, and security. It’s ideal for teams prioritising simplicity over granular control, particularly for production workloads. For advanced use cases requiring customisations, standard EKS with managed node groups remains a better fit.</p><h3>Resources</h3><ul><li><a href="https://docs.aws.amazon.com/eks/latest/userguide/eks-auto-mode.html">AWS EKS Auto Mode Documentation</a></li><li><a href="https://aws.amazon.com/blogs/containers/eks-auto-mode-security-best-practices/">Deep Dive on EKS Auto Mode Security</a></li></ul><h3>About the Writer ✍️</h3><p>Alex Li is a Cloud Architect passionate about demystifying AWS services. Follow for insights on cloud infrastructure and DevOps best practices.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=21bb27b34bd7" width="1" height="1" alt=""><hr><p><a href="https://towardsaws.com/simplifying-kubernetes-management-with-amazon-eks-auto-mode-21bb27b34bd7">Simplifying Kubernetes Management with Amazon EKS Auto Mode</a> was originally published in <a href="https://towardsaws.com/">Towards AWS</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AWS Sigv4 in 3 mins]]></title>
            <link>https://towardsaws.com/aws-sigv4-in-3-mins-c324d20f19cf?source=rss-47c0a821df07------2</link>
            <guid isPermaLink="false">https://medium.com/p/c324d20f19cf</guid>
            <category><![CDATA[security]]></category>
            <category><![CDATA[ami]]></category>
            <category><![CDATA[authentication]]></category>
            <category><![CDATA[identity]]></category>
            <category><![CDATA[aws]]></category>
            <dc:creator><![CDATA[Alex Z. Li]]></dc:creator>
            <pubDate>Thu, 24 Nov 2022 21:47:04 GMT</pubDate>
            <atom:updated>2023-01-12T20:06:16.518Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DWTjn9BNCNgZYTLcsLYrVg.png" /><figcaption>Sigv4 signing API call to AWS</figcaption></figure><h3>Context</h3><p>AWS uses <a href="https://docs.aws.amazon.com/general/latest/gr/signature-version-4.html">Sigv4</a> protocol for signing and authenticating every API requests against API Services, no matter its from the AWS Console, AWS Command line interface or via SDK (i.e. Python boto3 or other SDKs).</p><h3>Why?</h3><p>Sharing the long-term user credential (Access key id/and Key secret ) over public internet even through SSL/TLS can be dangerous, as customers’ credential should never appear in API requests directly. The Sigv4 provides the ability to scope down and sign the API requests to AWS services and provide additional secure control to verify the request.</p><h3>What’s Sigv4?</h3><blockquote>Signature Version 4 (SigV4) is the process to add authentication information to Amazon API requests sent by HTTP.</blockquote><p>With Sigv4, customer’s secrets never appear API requests directly, and with the Sigv4 design meaning every request is finely scoped down and signed with a token that is valid only for a given <strong>AWS service</strong>, in a given <strong>AWS region</strong>, <strong>on a given day</strong>. All of this is checked and verified at the authentication stage, before AWS tries to authorise a request.</p><h3>Key benefits</h3><ul><li>Don’t include the long term credentials directly in the API request</li><li>Signing key derived from AWS secret key, but scoped down to <strong>region</strong>, <strong>services</strong> and on <strong>a given day</strong>.</li><li>Version binding (API signing version) to prevent from downgrade attack.</li></ul><h3>Signing and verifying Process</h3><p><strong>How Signature Version 4 works at a high level</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*opLpcZx1aHOJpJZdCyP7Cw.jpeg" /><figcaption>high level workflow for Sigv4 signing and verification</figcaption></figure><ol><li>Create a canonical request.</li></ol><pre>CanonicalRequest =<br>  HTTPRequestMethod + &#39;\n&#39; +<br>  CanonicalURI + &#39;\n&#39; +<br>  CanonicalQueryString + &#39;\n&#39; +<br>  CanonicalHeaders + &#39;\n&#39; +<br>  SignedHeaders + &#39;\n&#39; +<br>  HexEncode(Hash(RequestPayload))</pre><p>2. Use the canonical request and additional metadata to create a string for signing.</p><pre>#Create a Canonical Request<br>CanonicalRequest =<br>  HTTPRequestMethod + &#39;\n&#39; +<br>  CanonicalURI + &#39;\n&#39; +<br>  CanonicalQueryString + &#39;\n&#39; +<br>  CanonicalHeaders + &#39;\n&#39; +<br>  SignedHeaders + &#39;\n&#39; +<br>  HexEncode(Hash(RequestPayload))</pre><p>3. Derive a signing key from your AWS secret access key. Then use the signing key, and the string from the previous step, to create a signature.</p><pre>#Create a string to sign <br>StringToSign =<br>    Algorithm + \n +<br>    RequestDateTime + \n +<br>    CredentialScope + \n +<br>    HashedCanonicalRequest</pre><pre>#Derive a signing key <br>kSecret = your secret access key<br>kDate = HMAC(&quot;AWS4&quot; + kSecret, Date)<br>kRegion = HMAC(kDate, Region)<br>kService = HMAC(kRegion, Service)<br>kSigning = HMAC(kService, &quot;aws4_request&quot;)</pre><pre>signature = HexEncode(HMAC(derived signing key, string to sign))</pre><p>4. Verification</p><p>When an AWS service receives the request, it performs the same steps that you did to calculate the signature you sent in your request. AWS then compares its calculated signature to the one you sent with the request. If the signatures match, the request is processed. If the signatures don’t match, the request is denied.</p><h3>Wrap up</h3><p>Normally, the Sigv4 signing process is abstracted away from users, if you are AWS Management Console, AWS CLI, or AWS SDKs. But if you are doing custom implementation, you can implement your own signing process based the AWS public documentation on Sigv4.</p><p>Now you know, how the Sigv4 works, and why AWS implement the singing protocol to protect your credential and verify the requests, so Go forth and stay secure in your AWS environment.</p><h3><strong>Resources</strong></h3><ul><li>How AWS SIGv4 and SIGv4A work — <a href="https://www.youtube.com/watch?v=BpTpcUA_F2w">https://www.youtube.com/watch?v=BpTpcUA_F2w</a> (recommend )</li><li>AWS Sigv4 document <a href="https://docs.aws.amazon.com/general/latest/gr/signature-version-4.html">https://docs.aws.amazon.com/general/latest/gr/signature-version-4.html</a></li><li>AWS SIGv4 and SIGv4A — <a href="https://shufflesharding.com/posts/aws-sigv4-and-sigv4a">https://shufflesharding.com/posts/aws-sigv4-and-sigv4a</a></li></ul><h4><strong>About Writer </strong>✍️</h4><p><a href="https://www.linkedin.com/in/alexzige/">Alex Li</a> is a Cloud Architect who writes about the AWS Cloud and Productivity tips.</p><p>Happy reading!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c324d20f19cf" width="1" height="1" alt=""><hr><p><a href="https://towardsaws.com/aws-sigv4-in-3-mins-c324d20f19cf">AWS Sigv4 in 3 mins</a> was originally published in <a href="https://towardsaws.com/">Towards AWS</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[1000 ft view- AWS Security]]></title>
            <description><![CDATA[<div class="medium-feed-item"><p class="medium-feed-image"><a href="https://medium.com/@Galaxy_Watcher/1000-ft-view-aws-security-4e3238744a69?source=rss-47c0a821df07------2"><img src="https://cdn-images-1.medium.com/max/600/0*bHbhPalvH4lCOM-y.png" width="600"></a></p><p class="medium-feed-snippet">Security is job Zero at AWS, we treat security seriously for both internal services and our customers and constantly raising security bar&#x2026;</p><p class="medium-feed-link"><a href="https://medium.com/@Galaxy_Watcher/1000-ft-view-aws-security-4e3238744a69?source=rss-47c0a821df07------2">Continue reading on Medium »</a></p></div>]]></description>
            <link>https://medium.com/@Galaxy_Watcher/1000-ft-view-aws-security-4e3238744a69?source=rss-47c0a821df07------2</link>
            <guid isPermaLink="false">https://medium.com/p/4e3238744a69</guid>
            <category><![CDATA[aws-certification]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[aws-cloud]]></category>
            <category><![CDATA[security]]></category>
            <category><![CDATA[aws]]></category>
            <dc:creator><![CDATA[Alex Z. Li]]></dc:creator>
            <pubDate>Thu, 17 Nov 2022 21:13:20 GMT</pubDate>
            <atom:updated>2022-11-17T21:13:20.132Z</atom:updated>
        </item>
        <item>
            <title><![CDATA[AWS Control Tower in 3 minutes]]></title>
            <link>https://towardsaws.com/aws-control-tower-in-3-minutes-cc90bc500499?source=rss-47c0a821df07------2</link>
            <guid isPermaLink="false">https://medium.com/p/cc90bc500499</guid>
            <category><![CDATA[security]]></category>
            <category><![CDATA[landingzone]]></category>
            <category><![CDATA[aws-control-tower]]></category>
            <category><![CDATA[cloud-computing]]></category>
            <category><![CDATA[aws]]></category>
            <dc:creator><![CDATA[Alex Z. Li]]></dc:creator>
            <pubDate>Tue, 07 Jun 2022 09:14:55 GMT</pubDate>
            <atom:updated>2022-11-22T10:08:55.480Z</atom:updated>
            <content:encoded><![CDATA[<p>To understand AWS Control Tower, first we need to explore the concepts of the landing zone in the cloud environment.</p><h4><strong>What’s a landing zone?</strong></h4><p>“Landing zone” in the cloud context refers to a secure and governed environment to launch your applications. Think of a launch pod that launch rockets like the diagram below 🚀</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*6Wx17n_D6eCqVE9S.jpg" /></figure><h4>Why it’s important?</h4><p>One of the biggest advantages for adopting to cloud is to lower the TCO (total cost of ownership), harnessing the value of the cloud native services like serverless, IOT and AI/ML, however in the traditional enterprise or government infrastructure environments, the organisation also needs to embrace security control, measurements and compliance and auditing, and often those needs to be take into consideration when comes to design your AWS or any other Cloud operating environments.</p><p>Landing zone helps addressing those by adopting a multi-account strategy and with orchestration services to help the organisation achieve higher security postures but also enable the organisation to innovate faster and unleash the true power of the public cloud platform.</p><p>Landing zone in AWS context refers to:</p><blockquote>The landing zone is a well-architected, multi-account baseline that follows AWS best practices<strong>.</strong></blockquote><h4>What is a Control Tower?</h4><p>Think of control tower helps create and managing your landing zone environment, give you the peace of mind operation your complex cloud environment.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*whMKkvty-3HqlV3i.png" /></figure><blockquote>Control Tower is managed services provided by AWS and offers the easiest way to set up and govern a secure, multi-account AWS environment. It establishes a landing zone that is based on best-practices blueprints, and it enables governance using guardrails you can choose from a pre-packaged list.</blockquote><p>In short, AWS Control Tower is a managed services that provides AWS best practices to operate in a multi-account environment, it boosts organisation with the agility and allows them to innovate fast with confidence.</p><h4>What’s in a AWS default control tower?</h4><p>A default control tower once you enabled in your AWS master account, it has:</p><ul><li>Setup the organisation and consolidate the billings</li><li>Configure federated login with AWS SSO</li><li>Enable default mandatory guardrails</li><li>Configure centralised log for CloudTrail for log achieve account</li><li>Configure Security/Audit account for security auditing</li></ul><p>and setup relevant services roles and service control policy to support the best practice multi-accounts orchestrations and governance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*wuMnxMasLtG08d_z.png" /><figcaption>Control Tower Enabled Multi-accounts AWS environment</figcaption></figure><p>Of cause, many enterprises and organisations often have its specially needs to manage their networks resource, where the base control tower setup cannot meet their needs, Control Tower has a solution <a href="https://aws.amazon.com/solutions/implementations/customizations-for-aws-control-tower/">Control Tower Customisation </a>to help design and customise your own control tower, which we will touch on further.</p><h4>Next up</h4><p>In the next chapters, we will deep dive into each of the components of the Control Tower, and explain hows things can be put together to address the particular needs from different teams within your organisation.</p><h4>Resources</h4><ul><li>AWS Control Tower: <a href="https://aws.amazon.com/controltower/">https://aws.amazon.com/controltower/</a></li><li>Control Tower best practice: <a href="https://docs.aws.amazon.com/controltower/latest/userguide/best-practices.html">https://docs.aws.amazon.com/controltower/latest/userguide/best-practices.html</a></li><li>Customisation for Control Tower: <a href="https://aws.amazon.com/solutions/implementations/customizations-for-aws-control-tower/">https://aws.amazon.com/solutions/implementations/customizations-for-aws-control-tower/</a></li></ul><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=cc90bc500499" width="1" height="1" alt=""><hr><p><a href="https://towardsaws.com/aws-control-tower-in-3-minutes-cc90bc500499">AWS Control Tower in 3 minutes</a> was originally published in <a href="https://towardsaws.com/">Towards AWS</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AWS 新工作漫谈]]></title>
            <link>https://medium.com/@Galaxy_Watcher/aws-%E6%96%B0%E5%B7%A5%E4%BD%9C%E6%BC%AB%E8%B0%88-ae299d3e66d8?source=rss-47c0a821df07------2</link>
            <guid isPermaLink="false">https://medium.com/p/ae299d3e66d8</guid>
            <dc:creator><![CDATA[Alex Z. Li]]></dc:creator>
            <pubDate>Mon, 09 May 2022 09:55:08 GMT</pubDate>
            <atom:updated>2022-05-09T09:56:28.786Z</atom:updated>
            <content:encoded><![CDATA[<p>一转眼来到AWS已经四个月了，这四个月以来，认识了不少新人，了解了些新公司的文化，学了一些新技能，我想主要谈谈我的一些感悟。</p><p><strong>亚马逊文化</strong></p><p>亚马逊倡导的是14个leadership principles 和 self serving 文化。 内部process非常松弛，希望员工能够做决定时可以以leadership principle 为核心。实现快速决策。这个以后可以说的东西很多，以后可以慢慢展开来说。</p><p><strong>“Let go with the ego”</strong></p><p>不得不说，亚麻的人都很厉害，L6以上的都至少15–20+ 年以上的经验。无论从阅历，技能和人脉都算是行业拔尖的那批人。如果你是个很有ego的人，来到AWS 很多人会一时间很难找到自己的位置。“let go with the ego” 拥有这growth mindset，把每一天看成学习的一部分。</p><p>记得，你从比较难打交道的客户上，学到的东西会最多，而且这些经验将会终身受用。</p><p>特别是我从app dev背景转过来有两大短板需要提升：</p><ul><li>AWS cloud 相关的技能：由于NZ 很多客户还在初级的cloud adoption，所以很多project都是和CAF, MAP和migration相关。一般对Cloud infrastructure 和 networking 需要比较深刻的理解。 其中对我的一个挑战是，短时间内完成对底层networking，和security的基础的建设。</li><li>Consulting excellence： 由于我们的工作有很多时候是和客户做workshop，consulting是我们工作的一大部分。作为专业AWS的consulting，我们需要从客户出发，利用我们的技术知识，去切实解决客户的痛点和难点，从而得到顾客的信任。</li></ul><p>这四个月以来，陆续完成了：</p><ul><li>AWS文化，工作方式，team有了些理解</li><li>对AWS一些底层服务有了100–200的知识储备</li><li>对AWS networking，hybrid connectivity有了稍微深入的理解</li><li>Deep dive了control tower 和landing zone -企业级 multi-accounts 策略上云基础</li><li>成功通过了AWS内部的AWesome builder 项目</li><li>成功的考过了三个associates level的认证：</li><li>AWS Solution Architect Associate</li><li>AWS Developer Associate</li><li>AWS Sysops Associate</li></ul><p>目前还在加紧学习VMware Cloud on AWS，希望能给新项目有帮助。</p><p>未来几个月，希望能拿下Security Speciality 的认证，再往AWS advance networking 冲击。</p><p>写给新人：</p><ul><li>有很多人问，有没有什么tips给新加入的新人，如下是我个人的感悟：</li></ul><p><strong>建立自己知识库</strong></p><ul><li>AWS 是个崇尚self-serving的公司，没有太多的hand holding，所以需要新人在onboarding时期可以很快建立自己的知识体系和知识库：其中包括去哪里寻找知识，哪些人可以寻求帮助，哪些slack可能会有类型的信息等等。</li></ul><p><strong>建立自己的人脉和社交网络</strong></p><ul><li>来了AWS最重要的一点是人脉的建立，人和人因为认识，产生信任，才能跟好的交流，找到自己的位置，需求导师的帮忙，可以帮你在短时间建立信心，积极的工作。</li></ul><p><strong>寻找自己的飞轮</strong></p><ul><li>AWS的信息量极大，如果你是一个好奇的人，你会发现对每个信息都不想放过，这样往往导致你不能很好的focus，精力消耗的过快，回头一看什么也没学到。</li><li>其实不如找到自己的抓手，静下心来努力深耕，在不断的深入中，开始做出一些成绩，建立自信，从而实现正向反馈，和飞轮效应。</li></ul><p>暂时就先写到这，与君共勉。</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ae299d3e66d8" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Management & leadership Series – things I’ve learnt the hard way]]></title>
            <link>https://medium.com/@Galaxy_Watcher/management-leadership-series-things-ive-learnt-in-a-hard-way-ccfcc0df7431?source=rss-47c0a821df07------2</link>
            <guid isPermaLink="false">https://medium.com/p/ccfcc0df7431</guid>
            <category><![CDATA[leadership-development]]></category>
            <category><![CDATA[management]]></category>
            <category><![CDATA[leadership]]></category>
            <category><![CDATA[coaching]]></category>
            <dc:creator><![CDATA[Alex Z. Li]]></dc:creator>
            <pubDate>Sat, 04 Dec 2021 22:14:40 GMT</pubDate>
            <atom:updated>2023-01-05T05:33:25.882Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/750/0*TvMTYkcNz4rrTZ6W.jpg" /></figure><blockquote>Wow, congrats!</blockquote><blockquote>You’ve got promoted to a new management role!</blockquote><p>You have made the first step towards your management journey, it’s all very shining and exciting, you have a few folks now reporting to you, and now you are responsible for helping them grow and thrive.</p><p>A good manager knows how to create a high performing team to thrive. Management is not top down with authority, the true leadership means genius cares and supports your team, and helping them thrive!</p><p>I would to like share some personal learnings that I learnt as a tech manager, and hope this could be a little helpful for you and help avoiding those mistakes, those are breaking down into the following categories, we will do a series of topics to cover those one by one.</p><ul><li><em>Delegation / empowerment</em></li><li><em>Time/energy management</em></li><li><em>Prioritisation</em></li><li><em>Earn trust</em></li><li><em>Feedbacks</em></li></ul><h4>Avoid the trap: Don’t try to do it all</h4><p>Most of the time, we get promoted because we have been good practitioners and excelled at our previous job level; we are good at solving problems (tech and non-tech), but as a manger or leader, it has more to it. From now on, you are measured against team outputs rather than individual. The hard thing is a mindset shift from <strong><em>do-it-your-own</em></strong> to <strong><em>empower</em></strong> others.</p><p><strong>Delegate &amp; empowerment</strong></p><p>You might heard the following from newly promoted manager or leads:</p><blockquote>“It might take me 10 mins to do rather than explain to someone for 2 hours”</blockquote><blockquote>“This project requires a lot of context, is it probably better for me to do it this time?”</blockquote><blockquote>“I’m not sure he/she can complete it by the deadline for the quality that i want to see, is it safer for me just to do it?”</blockquote><p>As a new manager, we have all been there before, it always seems easier for us to jump in and work on the tasks, because</p><blockquote>1. Perhaps it’s more efficient for us to just do it;</blockquote><blockquote>2. We could control the outcomes to our satisfaction, that remove any guesswork;</blockquote><p>However, if you cannot get over this way of thinkings, it could potentially create a lot of burden on you, and as a results:</p><blockquote>1. You will have no time for focus on more strategic work;</blockquote><blockquote>2. You will take away the autonomy out of your team, your team members will be less empowered to make decisions and taking ownership;</blockquote><p>Over a period of time, you will be eventually burnout and your team may feel less empowered and have less motivation!</p><h4>Tips for leaders and managers</h4><p>For those new leads and managers, here are some practice tips you might find helpful.</p><blockquote>Accept people might not do as good as you do at the first go, if they can do 70%,<strong> that’s okay!</strong></blockquote><p><strong>Professionalism is your biggest enemy</strong></p><p>The No.1 tip for all new managers is to change mindset and recognise that if the work is done in 70%, that’s okay for them on the first try! Be clear with areas of improvements, asking for their solutions and allow them to take ownership, get them to own their mistakes.</p><blockquote>Don’t jump in too early, give the space, your job is to ensure the end results, not the series of tasks.</blockquote><p><strong>Jump in to help? No, you should wait!</strong></p><p>We hire the best and develop most of our people, why not trust them to do the best work? Though they might not work 100% the same way you’d like to see, but if the end results are the same, why we need to correct the course? Of course, you will help them when they are facing huge challenges or obstacles and ask for help, but do not jump in too early, if they haven’t ask for, if you do, you take away the autonomy and the responsibility is back to you again!</p><p>Reflecting the learnings, when things fail, remember it’s never been a blame game, try to evolve and learn from the failure and cultivate a strong safety net.</p><p>That’s it for now, stay tuned for the next one when we talk about “time and energy management”.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ccfcc0df7431" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Web, native, and hybrid apps — how to choose the right tech stack]]></title>
            <link>https://medium.com/@Galaxy_Watcher/web-native-and-hybrid-apps-how-to-choose-the-right-tech-stack-51485d7c732f?source=rss-47c0a821df07------2</link>
            <guid isPermaLink="false">https://medium.com/p/51485d7c732f</guid>
            <category><![CDATA[native-app]]></category>
            <category><![CDATA[mobile-apps]]></category>
            <category><![CDATA[mobility]]></category>
            <category><![CDATA[cross-platform]]></category>
            <category><![CDATA[web-app-development]]></category>
            <dc:creator><![CDATA[Alex Z. Li]]></dc:creator>
            <pubDate>Sun, 18 Jul 2021 23:06:53 GMT</pubDate>
            <atom:updated>2021-07-18T23:06:53.787Z</atom:updated>
            <content:encoded><![CDATA[<h3>Web, native, and hybrid apps — how to choose the right tech stack</h3><p>Choose the right tech stack for your business by understanding the pros and cons of web, native, and hybrid apps.</p><p><em>(Originally published at </em><a href="https://www.springload.co.nz/blog/web-native-and-hybrid-apps-how-choose-right-tech-stack/"><em>https://www.springload.co.nz</em></a><em>)</em></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/960/0*AclTgc_3BGHH18NO.jpg" /></figure><p>You have an idea for a mobile app, what’s next? One of the most important decisions you’ll need to make is choosing whether to build a web, native, or hybrid app. Each approach has different pros and cons, and these need to be considered in the context of your requirements, budget, and users’ needs.</p><h3>What is a web app?</h3><h4>Responsive web apps</h4><p>Responsive web apps work across desktop and mobile browsers. Out of the three approaches, this is the most straightforward and cost-effective option, because there is one codebase for all platforms. Responsive design can make a web app look different depending on the device used to view it and optimised for a mobile experience.</p><h4>Progressive web apps</h4><p>Progressive web apps (PWAs) use web capabilities to offer users an app-like experience similar to a native mobile app. Progressive web apps are relatively new to web app development and support a variety of native capabilities such as working offline, limited push notifications, and access to hardware like the camera of a device.</p><h3>What is a hybrid app?</h3><p>A hybrid app is a web app that’s packaged into a native app shell. Hybrid apps are a good solution if you already have a web app, but also want to offer a mobile experience that can access some native hardware features.</p><p>Whilst hybrid apps are faster to develop, they’re not suited for performance-intensive tasks such as on-device machine learning, augmented reality (AR), and virtual reality (VR) support.</p><h3>What is a native app?</h3><h4>Native apps</h4><p>Unlike a web or hybrid app, a native app is built for specific platforms (iOS or Android) using official platform frameworks and SDKs (Software development kit). Native apps can provide a fully customisable and convenient experience for users and full access to device hardware, however, they require more time and resources to develop.</p><p>If your service relies on using hardware such as GPS, Bluetooth, or the camera of a device, native apps will give you the best integration.</p><h4>Cross-platform apps</h4><p>Cross-platform apps are a cost-effective solution for developing native apps. A cross-platform app uses a single codebase across multiple platforms which reduces the cost of development and maintenance. For your users, cross-platform and native apps offer a similar experience: both are installed and updated from the App Store or Google Play.</p><p>Cross-platform development works best for apps that don’t require lots of hardware features and aren’t performance-sensitive. Developing and maintaining a cross-platform app requires fewer resources than a native app, and uses frameworks such as Flutter, React Native, or Xamarin.</p><h3>Which approach is right for my business and users?</h3><p>Mobile and web apps aren’t mutually exclusive. A lot of products use both platforms to cater for different users’ needs, but if you’re just starting your app journey, we recommend prioritising one approach first.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*kgmVeCU4N-n0HGlg.jpg" /></figure><h3>Next?</h3><p>interested to find out more? Drop us a line.</p><p><em>Originally published at </em><a href="https://www.springload.co.nz/blog/web-native-and-hybrid-apps-how-choose-right-tech-stack/"><em>https://www.springload.co.nz</em></a><em>.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=51485d7c732f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[IT job hunting in New Zeland]]></title>
            <link>https://medium.com/chinese-it-association-nz/it-job-hunting-in-new-zeland-991780d6d0e7?source=rss-47c0a821df07------2</link>
            <guid isPermaLink="false">https://medium.com/p/991780d6d0e7</guid>
            <category><![CDATA[meetup]]></category>
            <category><![CDATA[featured]]></category>
            <dc:creator><![CDATA[Alex Z. Li]]></dc:creator>
            <pubDate>Thu, 16 May 2019 10:29:45 GMT</pubDate>
            <atom:updated>2019-06-15T09:47:47.019Z</atom:updated>
            <content:encoded><![CDATA[<h3>Job Hunting in New Zealand</h3><h4>Top tips from Wellington IT recruiter — Adelaide</h4><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2HgcYUuy-J9pN8BATk1WZQ.jpeg" /></figure><p><strong>How to secure a dream job in NZ?</strong></p><p>We have been asked a lot about how to write a CV and prepare for interviews. Adelaide from <a href="https://www.findrecruitment.co.nz/">FindiT</a> shed some lights on the top tips for securing a dream IT job in New Zealand in our meet-ups.</p><p>During the presentation, Adelaide covered the common mistakes of CV writing, a typical interview process and highlighted a few key areas for our members to focus in order to increase the chance of nailing a job offer.</p><h4>A good CV consists of</h4><ul><li><strong>Name</strong> — include English name or pronunciation of your Chinese name</li><li><strong>Intro</strong>/personal statement</li><li><strong>Visa</strong> <strong>status</strong></li><li><strong>Strengths</strong> &amp; <strong>Overview of</strong> <strong>technical skills</strong></li><li><strong>Experience</strong></li><li><strong>Education</strong></li><li><strong>Hobbies</strong></li></ul><p>Note*: (3–4 pages, White space, use of bullet points, easy to scan &amp; read, no long paragraphs)</p><h4>A good experience session should highlight</h4><ul><li><strong>Job Title </strong>— period</li><li><strong>Company</strong> (if your previous company is outside NZ, it would be a good idea to include some intro, so people can easily conceptualise it.)</li><li><strong>Role &amp; Responsibilities (</strong>list the key roles and responsibilities, help to sell your skills and experience better.)</li><li><strong>Achievements</strong> (if there’s any)</li><li><strong>Technology adopted</strong></li></ul><h3><strong>Key takeaways</strong></h3><ul><li>Put yourself in other&#39;s shoes; they may not have the context of your previous experience and achievements; it’s your job to sell your story.</li><li>Practice the content in your CV;</li><li>Be confident and genuine;</li><li>Get used to chit-chat; (i.e weather and coffees)</li><li>Get used to the coffee culture ;)</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*yfGKg0sDRYF0FgtSdypqVQ.jpeg" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*loQr3LU2vg4PsUqTmyqVdA.jpeg" /></figure><h4>For more information</h4><p><a href="https://docs.google.com/presentation/d/1zWQN73l4ytcv09kyGe0Na6v5l-FNnQ61OOo1WjByBmM/edit#slide=id.p5">Presentation slides can be viewed here</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=991780d6d0e7" width="1" height="1" alt=""><hr><p><a href="https://medium.com/chinese-it-association-nz/it-job-hunting-in-new-zeland-991780d6d0e7">IT job hunting in New Zeland</a> was originally published in <a href="https://medium.com/chinese-it-association-nz">Chinese IT Association NZ</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[CITANZ Member end of year Carnival]]></title>
            <link>https://medium.com/chinese-it-association-nz/citanz-member-end-of-year-carnival-b88dbf2fd640?source=rss-47c0a821df07------2</link>
            <guid isPermaLink="false">https://medium.com/p/b88dbf2fd640</guid>
            <category><![CDATA[events]]></category>
            <category><![CDATA[featured]]></category>
            <dc:creator><![CDATA[Alex Z. Li]]></dc:creator>
            <pubDate>Tue, 23 Apr 2019 09:54:15 GMT</pubDate>
            <atom:updated>2019-06-15T10:53:28.674Z</atom:updated>
            <content:encoded><![CDATA[<h3>CITANZ Event End-of-Year Carnival</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aEcmELrQb0tsxE_qmc1tkA.jpeg" /></figure><h3>Our Member Carnival</h3><p>On 27th Jan, we had our very first member carnival with our first batch of members and their extended family!</p><p>The event has been broken into a couple of sections:</p><ul><li>End of year CITANZ announcements by the president</li><li>Awards ceremony for our guest speakers and volunteers</li><li>Member introduction</li><li>BBQ/Lunch</li><li>GAMES</li></ul><p>This is the first time for CITNAZ to organise an event like this scale, we all had a great time with our members and we shared lots of stories.</p><p>A huge shout-out to all our members and our lovely volunteers, it won’t be possible without your support!</p><p>Here are a video teaser and some great photos:</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fwww.youtube.com%2Fembed%2FYLhGVSvXSM8%3Ffeature%3Doembed&amp;url=http%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3DYLhGVSvXSM8&amp;image=https%3A%2F%2Fi.ytimg.com%2Fvi%2FYLhGVSvXSM8%2Fhqdefault.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=youtube" width="854" height="480" frameborder="0" scrolling="no"><a href="https://medium.com/media/5d6905b6a5abe70eff17e0aac39cd2b0/href">https://medium.com/media/5d6905b6a5abe70eff17e0aac39cd2b0/href</a></iframe><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*-pBsU-sNm4sR9G2HtCD4aw.jpeg" /><figcaption>Our lovely CITANZ ladies</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*upN3NWg0sWs6sbW4kGV-Sw.jpeg" /><figcaption>Social events and games</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qTZzJrVGGzn5CE55l2JsQg.jpeg" /><figcaption>Socialising with members and families</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wIJ1Z1hYdt-SCwFSsVTvOQ.jpeg" /><figcaption>Special thanks to our CITANZ volunteers</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*FyfAnJ8AE8ODuOfAmpcLDg.jpeg" /><figcaption>Food ;)</figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*qznnr8bPuggXZtGmkis1Jw.jpeg" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=b88dbf2fd640" width="1" height="1" alt=""><hr><p><a href="https://medium.com/chinese-it-association-nz/citanz-member-end-of-year-carnival-b88dbf2fd640">CITANZ Member end of year Carnival</a> was originally published in <a href="https://medium.com/chinese-it-association-nz">Chinese IT Association NZ</a> on Medium, where people are continuing the conversation by highlighting and responding to this story.</p>]]></content:encoded>
        </item>
    </channel>
</rss>