<?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 Serkan SAKINMAZ on Medium]]></title>
        <description><![CDATA[Stories by Serkan SAKINMAZ on Medium]]></description>
        <link>https://medium.com/@serkansakinmaz?source=rss-583460cf9605------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*g6IXzG-rOBCe0p39.jpg</url>
            <title>Stories by Serkan SAKINMAZ on Medium</title>
            <link>https://medium.com/@serkansakinmaz?source=rss-583460cf9605------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 00:19:02 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@serkansakinmaz/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 to create Databricks with Terraform]]></title>
            <link>https://serkansakinmaz.medium.com/how-to-create-databricks-with-terraform-c1f0bc2a1daa?source=rss-583460cf9605------2</link>
            <guid isPermaLink="false">https://medium.com/p/c1f0bc2a1daa</guid>
            <category><![CDATA[azure-databricks]]></category>
            <category><![CDATA[terraform]]></category>
            <category><![CDATA[deployment]]></category>
            <dc:creator><![CDATA[Serkan SAKINMAZ]]></dc:creator>
            <pubDate>Wed, 23 Jul 2025 09:48:54 GMT</pubDate>
            <atom:updated>2025-07-23T09:48:54.440Z</atom:updated>
            <content:encoded><![CDATA[<p>In this blog post, I’ll guide you through deploying <strong>Azure Databricks</strong> using <strong>Terraform</strong>. It’s a straightforward process.Let’s walk through it step by step.</p><h3>Step-by-Step Guide</h3><p><strong>1. Create a Terraform project folder</strong></p><p>Start by creating a directory to hold your Terraform configuration files:</p><pre>mkdir terraform-databricks<br>cd terraform-databricks</pre><p><strong>2. Create a </strong><strong>main.tf file</strong></p><p>Inside your folder, create a main.tf file. This is where your Terraform configuration will live.</p><p><strong>3. Add your Terraform configuration</strong></p><p>Paste the appropriate Terraform code for provisioning Azure Databricks in the main.tf file. (You can adjust this based on your resource group, region, and other settings.)</p><pre>terraform {<br>  required_version = &quot;&gt;=1.3.0&quot;<br>  required_providers {<br>    azurerm = {<br>      source = &quot;hashicorp/azurerm&quot;<br>      version = &quot;&gt;=3.111.0&quot;<br>    }<br>  }<br>}<br><br>provider &quot;azurerm&quot; {<br>  features {}<br>  skip_provider_registration = true<br>  subscription_id = &quot;bf12bd6a-ea3e-4352-9d57-500f3db25728&quot;<br>}<br><br>resource &quot;azurerm_resource_group&quot; &quot;this&quot; {<br>  name = &quot;deploydb&quot;<br>  location = &quot;Germany West Central&quot;<br><br>}<br><br>resource &quot;azurerm_databricks_workspace&quot; &quot;this&quot; {<br>  name = &quot;nc-serkan-db&quot;<br>  location = azurerm_resource_group.this.location<br>  resource_group_name = azurerm_resource_group.this.name<br>  managed_resource_group_name = join(&quot;&quot;,[azurerm_resource_group.this.name,&quot;_mrg&quot;])<br>  sku = &quot;premium&quot;<br><br>}<br><br>output &quot;databricks_host&quot; {<br>  value = &quot;https://${azurerm_databricks_workspace.this.workspace_url}&quot;<br>}</pre><p><strong>4. Authenticate with Azure</strong></p><p>Use the Azure CLI to sign in:</p><pre>az login</pre><p><strong>5. Initialize the Terraform project</strong></p><p>This sets up the required Terraform providers and backend configuration:</p><pre>terraform init</pre><p><strong>6. Apply the configuration</strong></p><p>Run the following to deploy the infrastructure:</p><pre>terraform apply</pre><p>Confirm the action when prompted. Once the process is complete, your <strong>Azure Databricks workspace</strong> will be live and ready to use.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IjL3Sbwf60evRcEg5km5VA.png" /></figure><p><strong>7. (Optional) Tear down the infrastructure</strong></p><pre>terraform destroy --auto-approve</pre><h3>Final Thoughts</h3><p>Using Terraform to provision Azure Databricks is a powerful way to manage infrastructure as code. It enables reproducibility, version control, and automation,perfect for cloud engineering and DevOps workflows. Give it a try and streamline your cloud deployments!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c1f0bc2a1daa" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Analyze text with Azure AI Foundry]]></title>
            <link>https://serkansakinmaz.medium.com/analyze-text-with-azure-ai-foundry-2f15c8ca0f09?source=rss-583460cf9605------2</link>
            <guid isPermaLink="false">https://medium.com/p/2f15c8ca0f09</guid>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[nlp]]></category>
            <category><![CDATA[extract-information]]></category>
            <category><![CDATA[azureaifoundry]]></category>
            <dc:creator><![CDATA[Serkan SAKINMAZ]]></dc:creator>
            <pubDate>Fri, 11 Jul 2025 14:17:44 GMT</pubDate>
            <atom:updated>2025-07-11T14:17:44.696Z</atom:updated>
            <content:encoded><![CDATA[<p>IIn this blog post, I’ll walk you through how to analyze text using <strong>Microsoft Azure AI Foundry</strong> — a powerful tool that enables you to extract structured information like cities, countries, and dates from unstructured text.</p><p>This tutorial is perfect for anyone looking to get started with Azure’s AI capabilities, especially for natural language processing tasks such as <strong>named entity recognition</strong>.</p><h3>What You’ll Learn</h3><ul><li>How to set up a project in Azure AI Foundry</li><li>How to use the <strong>Language Playground</strong></li><li>How to extract named entities (like city names, countries, and dates) from any text</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*hfN9jjKcMZaZ2a5VlK5d1Q.png" /></figure><h3>Step-by-Step Guide</h3><h3>Step 1: Open Azure AI</h3><p>Go to <a href="https://ai.azure.com/">https://ai.azure.com</a>.</p><h3>Step 2: Create a New Resource</h3><p>Navigate to the <strong>Management Center</strong> at<br> <a href="https://ai.azure.com/managementCenter/allResources">https://ai.azure.com/managementCenter/allResources</a><br> Click <strong>Create new</strong> to start a new project.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*vjoCpvSknXQ8yxLDsOtDHw.png" /></figure><h3>Step 3: Select Resource Type</h3><p>Choose <strong>Azure AI Foundry Resource</strong> and click <strong>Next</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xwg_yK0SCwj0mIRQLgBzhQ.png" /></figure><h3>Step 4: Configure Your Project</h3><p>Fill in the required details like project name, region, and resource group. Once completed, click <strong>Create</strong>.</p><blockquote>Note: It may take a few minutes for the project to be fully provisioned.</blockquote><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DAVV6DkTu1CMr8viTNp2Pg.png" /></figure><h3>Step 5: Access the Playground</h3><p>Once your project is ready, on the left-hand menu, click <strong>Playgrounds</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*iWtj2LyqaDNCiFCTITdG9g.png" /></figure><h3>Step 6: Open the Language Playground</h3><p>Find and select the <strong>Language Playground</strong> from the list of available tools.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_zqcOFSDKtf4BUNG7QDlMg.png" /></figure><h3>Step 7: Extract Named Entities</h3><p>In the Language Playground, select:</p><ul><li><strong>“Extract Information”</strong></li><li>Then choose <strong>“Extract Named Entities”</strong></li></ul><p>This module allows you to automatically identify and categorize entities like cities, dates, organizations, and more.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*dCkopmI-zTV35Qtbyw8F6Q.png" /></figure><h3>Example: Try It Yourself</h3><p>You can use your own text, or copy and paste the following sample:</p><pre>Where To Stay in Nuremberg (Nürnberg) in 2025<br>My Germany Vacation<br>https://www.mygermanyvacation.com › Blog<br>8 Best Hotels in Nuremberg&#39;s Old Town · 1. Where We Stayed: Gideon Hotel · 2. Large Family Rooms: Holiday Inn Nuremberg City Centre · 3. In the Central Market ...</pre><p>Paste this into the text editor within the Extract Named Entities module and click <strong>Run</strong>.</p><p>You’ll see Azure automatically categorize various parts of the text, identifying elements such as:</p><ul><li><strong>Cities</strong>: Nuremberg, Nürnberg</li><li><strong>Country</strong>: Germany</li><li><strong>Date</strong>: 2025</li><li><strong>Organizations/Locations</strong>: Gideon Hotel, Holiday Inn</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZOlBJlLh_QsuNWtpVypvjA.png" /></figure><h3>Final Thoughts</h3><p>Azure AI Foundry makes it incredibly easy to derive structured insights from raw text — no code required. Whether you’re working on travel blogs, business documents, or news articles, the platform helps you uncover the meaningful entities within your data.</p><p>Feel free to experiment with different types of content — you’ll be surprised how much information Azure can uncover!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2f15c8ca0f09" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Getting Started with Generative AI in Azure]]></title>
            <link>https://serkansakinmaz.medium.com/getting-started-with-generative-ai-in-azure-a5abebcab5ad?source=rss-583460cf9605------2</link>
            <guid isPermaLink="false">https://medium.com/p/a5abebcab5ad</guid>
            <category><![CDATA[azure-ai]]></category>
            <category><![CDATA[gpt-4]]></category>
            <category><![CDATA[azure-generative-ai]]></category>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[chatplayground]]></category>
            <dc:creator><![CDATA[Serkan SAKINMAZ]]></dc:creator>
            <pubDate>Thu, 10 Jul 2025 07:26:28 GMT</pubDate>
            <atom:updated>2025-07-10T07:26:28.217Z</atom:updated>
            <content:encoded><![CDATA[<p>In this blog post, I’ll walk you through how to use <strong>Generative AI</strong> on <strong>Microsoft Azure</strong>. Whether you’re a developer experimenting with AI or a business exploring intelligent solutions, this guide will help you get started using <strong>Azure AI Foundry.</strong></p><h3>🔹 Step-by-Step Guide to Using Generative AI in Azure</h3><h3>✅ Step 1: Open the Azure AI Portal</h3><p>Navigate to the Azure AI portal:<br> 🔗 <a href="https://ai.azure.com/">https://ai.azure.com</a></p><h3>✅ Step 2: Create an AI Resource</h3><p>Go to the resource management page:<br> 🔗 <a href="https://ai.azure.com/managementCenter/allResources">https://ai.azure.com/managementCenter/allResources</a><br> Click <strong>“Create”</strong> and select the <strong>Azure AI Foundry</strong> resource.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*b2Rd0vhm6XmLx4HBLi_ROg.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*T1cg-K4vb3Knja64jjjXQA.png" /></figure><h3>✅ Step 3: Fill in Resource Details</h3><p>Provide the necessary configuration:</p><ul><li><strong>Subscription</strong></li><li><strong>Resource Group</strong></li><li><strong>Region</strong></li><li><strong>Azure AI Foundry Resource Name</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*T1cg-K4vb3Knja64jjjXQA.png" /></figure><p>Provisioning might take a couple of minutes. Once it’s done, you’ll be directed to the <strong>Overview</strong> page where you’ll find important details such as:</p><ul><li>API Key</li><li>Agents</li><li>Templates</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*x22giP66q4PqFRgkVuq34w.png" /></figure><h3>✅ Step 4: Open the Playground</h3><p>From the left sidebar, select <strong>“Playgrounds”</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*mi69IvDV3HEzvfyBE1HMrg.png" /></figure><h3>✅ Step 5: Choose Chat Playground</h3><p>In the playgrounds section, click on <strong>“Chat Playground”</strong> to start interacting with the model.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*t76t3-ZMkAIfVwR7dtn7Kw.png" /></figure><h3>✅ Step 6: Create a Deployment</h3><p>Click <strong>“Create Deployment”</strong>, choose <strong>GPT-4o</strong>, confirm your selection, and deploy. This sets up your model for use.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3FTAtTI0fEfLbL83rZKmVw.png" /></figure><h3>✅ Step 7: Start Chatting!</h3><p>Once your deployment is live, you’ll see an interactive chat interface.</p><p>🎉 <strong>Great!</strong> You now have access to your own Generative AI chat model in Azure.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Nsx92rz7C9BwV3bI4czV2g.png" /></figure><p>✅ Step 8: Run Some Tests</p><p>Try entering prompts in the chat box — ask questions, generate text, or explore creative ideas.</p><p>🎯 <strong>Success!</strong> You’ve now experienced how Generative AI works with Azure AI Foundry.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*42k0W9_o_hhHu19JlMxzCQ.png" /></figure><h3>🧹 Next Steps &amp; Clean-Up</h3><p>When you’re done testing, don’t forget to <strong>delete your resources</strong> if you no longer need them. This helps avoid unnecessary costs and keeps your Azure environment tidy.</p><h3>🚀 Final Thoughts</h3><p>Generative AI is revolutionizing how we build and interact with software. With Azure AI Foundry, Microsoft makes it easy to explore powerful models like GPT-4o in just a few clicks.</p><p>Stay tuned for more posts where I’ll show you how to integrate these capabilities into your apps and workflows!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a5abebcab5ad" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Predicting Bicycle Rentals with Azure Machine Learning-Part 3]]></title>
            <link>https://serkansakinmaz.medium.com/predicting-bicycle-rentals-with-azure-machine-learning-part-3-500c20c38643?source=rss-583460cf9605------2</link>
            <guid isPermaLink="false">https://medium.com/p/500c20c38643</guid>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[azure-machine-learning]]></category>
            <category><![CDATA[azure-deploy]]></category>
            <category><![CDATA[predictions]]></category>
            <category><![CDATA[regression]]></category>
            <dc:creator><![CDATA[Serkan SAKINMAZ]]></dc:creator>
            <pubDate>Wed, 09 Jul 2025 08:43:40 GMT</pubDate>
            <atom:updated>2025-07-09T08:43:40.484Z</atom:updated>
            <content:encoded><![CDATA[<p>Welcome to Part 3 of our series on predicting bicycle rentals using Azure Machine Learning. In this part, we’ll walk through the steps to test the deployed model and interpret its predictions. If you haven’t already, make sure to check out [<a href="https://medium.com/@serkansakinmaz/predicting-bicycle-rentals-with-azure-machine-learning-part-2-26cf1344d61f">Part 2</a>] to set up your environment and deploy the model.</p><h3>🚴 Step-by-Step: Testing the Model</h3><h4>Step 1 — Select the Endpoint</h4><p>Navigate to the <strong>Endpoints</strong> section in Azure Machine Learning Studio. Choose the endpoint you deployed in the previous part of this series.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YuT1vP6OStQDyR_QS-tHHA.png" /></figure><h4>Step 2 — Open the Test Tab</h4><p>Click on the <strong>Test</strong> tab in the endpoint view. This tab allows you to send test data to your model and view the predictions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Fl8pemtu6FbqDS1yshzQpQ.png" /></figure><h4>Step 3 — Prepare Your Input</h4><p>Use the following JSON template as a starting point. Replace the values with your own input data to simulate different rental conditions.</p><pre>   {<br>  &quot;input_data&quot;: {<br>    &quot;columns&quot;: [<br>      &quot;day&quot;,<br>      &quot;mnth&quot;,<br>      &quot;year&quot;,<br>      &quot;season&quot;,<br>      &quot;holiday&quot;,<br>      &quot;weekday&quot;,<br>      &quot;workingday&quot;,<br>      &quot;weathersit&quot;,<br>      &quot;temp&quot;,<br>      &quot;atemp&quot;,<br>      &quot;hum&quot;,<br>      &quot;windspeed&quot;<br>    ],<br>    &quot;index&quot;: [0],<br>    &quot;data&quot;: [[1,1,2022,2,0,1,1,2,0.3,0.3,0.3,0.3]]<br>  }<br> }</pre><h4>Step 4 — Run the Test</h4><p>Click the <strong>Test</strong> button to send your data to the model. After a few moments, the prediction result will appear.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/676/1*8LXD0Wzf0LD3Duf83l7-0g.png" /></figure><p>🎉 <strong>Success!</strong><br> Your model has returned the predicted number of bicycle rentals for the given day. It’s a great way to gain insights into expected demand based on weather and calendar conditions.</p><p>💡 <strong>Tip:</strong> Try testing with different input values to see how season, temperature, or holidays affect rental numbers.</p><h3>🧹 Next Steps: Clean Up Resources</h3><p>Once you’re done testing, it’s a good idea to clean up your Azure resources to avoid unnecessary charges. You can delete the endpoint and other associated resources from the Azure portal.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=500c20c38643" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Predicting Bicycle Rentals with Azure Machine Learning-Part 2]]></title>
            <link>https://serkansakinmaz.medium.com/predicting-bicycle-rentals-with-azure-machine-learning-part-2-26cf1344d61f?source=rss-583460cf9605------2</link>
            <guid isPermaLink="false">https://medium.com/p/26cf1344d61f</guid>
            <category><![CDATA[azure-machine-learning]]></category>
            <category><![CDATA[azure-endpoints]]></category>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[regression]]></category>
            <category><![CDATA[predictions]]></category>
            <dc:creator><![CDATA[Serkan SAKINMAZ]]></dc:creator>
            <pubDate>Wed, 09 Jul 2025 08:32:26 GMT</pubDate>
            <atom:updated>2025-07-09T09:34:08.480Z</atom:updated>
            <content:encoded><![CDATA[<p>Welcome to Part 2 of our series on building a Bicycle Rental Prediction System. In this part, we’ll walk through the steps to deploy our trained model and create a real-time inference endpoint using Azure Machine Learning.</p><blockquote><em>📌 </em>If you’re just joining us, be sure to check out [<a href="https://medium.com/@serkansakinmaz/predicting-bicycle-rentals-with-azure-machine-learning-part-1-1ee10199c10c">Part 1</a>] where we trained the model.</blockquote><h3>🛠 Step 1 — Select the Completed Job</h3><p>Navigate to your list of completed jobs in Azure Machine Learning Studio.<br> From here, select the job that contains the model you want to deploy.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*PNP1GtT2HNG7Spv_u-Hb4Q.png" /></figure><h3>🔍 Step 2 — Choose the Best Model</h3><p>Go to the <strong>Overview</strong> tab of the selected job. Under the <strong>Best Model Summary</strong>, click on the algorithm name to view and access the best-performing model from your training run.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*0hoqulsYZj1JjcNuDi85WQ.png" /></figure><h3>🚀 Step 3 — Deploy the Model</h3><p>Click on the <strong>Deploy</strong> button, then select <strong>Real-time Endpoint</strong>.<br> This will open the configuration panel where you’ll set up the deployment environment.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*uJ2As91-Td0cbUKfYFVmlQ.png" /></figure><h3>⚙️ Step 4 — Configure Deployment Settings</h3><p>Provide the following details:</p><ul><li><strong>Deployment Name</strong>: Give your deployment a meaningful name (e.g., bike-rental-predictor).</li><li><strong>Instance Count</strong>: Typically set to 1 for basic testing; scale as needed for production.</li><li><strong>Virtual Machine Size</strong>: Choose based on your workload. For example, Standard_DS2_v2 is often sufficient for testing purposes.</li></ul><blockquote><em>⏳ </em>Deployment may take 5–10 minutes to complete.</blockquote><h3>🧯 Troubleshooting Note</h3><p>During deployment, you might encounter the following error:</p><pre>vbnet</pre><pre>CopyEdit</pre><pre>ResourceOperationFailure: Resource provider [N/A] isn&#39;t registered with Subscription [N/A].</pre><p><strong>Solution</strong>:<br> To resolve this, register the necessary resource providers:</p><ol><li>Go to <strong>Azure Portal</strong>.</li><li>Navigate to <strong>Subscriptions</strong> &gt; your subscription &gt; <strong>Resource Providers</strong>.</li><li>Search for and register:</li></ol><ul><li>Microsoft.PolicyInsights</li><li>Microsoft.Cdn</li></ul><p>You can also refer to the official <a href="https://aka.ms/register-resource-provider">troubleshooting guide</a> for additional help.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*Jbxi6d-FmI7oVt8Z63QnRQ.png" /></figure><h3>✅ Step 5 — Verify Deployment</h3><p>After a few minutes, your deployment should be complete.<br> To verify:</p><ul><li>Go to the <strong>Endpoints</strong> tab.</li><li>Check the <strong>Status</strong> to confirm it’s Healthy.</li><li>Copy the <strong>Endpoint URL</strong> — you’ll use this in the next step to make predictions.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*NZua2O8KYX1v3B30rJyNWQ.png" /></figure><h3>🎉 Wrapping Up</h3><p>In this part, we successfully deployed our machine learning model as a real-time endpoint. You’re now one step closer to serving predictions in a live application!</p><p>👉 <strong>Next Up</strong>: In [<a href="https://medium.com/@serkansakinmaz/predicting-bicycle-rentals-with-azure-machine-learning-part-3-500c20c38643">Part 3</a>], we’ll test the deployed endpoint by sending sample data and analysing the responses.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=26cf1344d61f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Predicting Bicycle Rentals with Azure Machine Learning — Part 1]]></title>
            <link>https://serkansakinmaz.medium.com/predicting-bicycle-rentals-with-azure-machine-learning-part-1-1ee10199c10c?source=rss-583460cf9605------2</link>
            <guid isPermaLink="false">https://medium.com/p/1ee10199c10c</guid>
            <category><![CDATA[regression]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[azure]]></category>
            <category><![CDATA[azure-machine-learning]]></category>
            <dc:creator><![CDATA[Serkan SAKINMAZ]]></dc:creator>
            <pubDate>Wed, 09 Jul 2025 07:23:47 GMT</pubDate>
            <atom:updated>2025-07-09T07:23:47.893Z</atom:updated>
            <content:encoded><![CDATA[<h3>Predicting Bicycle Rentals with Azure Machine Learning — Part 1</h3><p>In this post, we’ll walk through how to predict the number of bicycle rentals using <strong>Azure Machine Learning Service</strong>. By the end of this guide, you’ll have trained a machine learning model using automated machine learning (AutoML) in Azure, setting the foundation for deploying it in Part 2.</p><h3>Step-by-Step Guide to Building a Bicycle Rental Prediction Model</h3><h3>Step 1 — Create a Machine Learning Resource</h3><p>Start by navigating to the <strong>Azure Portal</strong>, and create a new <strong>Machine Learning</strong> resource from the available Azure services.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/470/1*MZKFpBIx9M54sBNNMn91pw.png" /></figure><h3>Step 2 — Set Up Your Workspace</h3><p>Once the resource is created, set up a <strong>workspace</strong>. You can safely accept the default options for:</p><ul><li><strong>Storage account</strong></li><li><strong>Key vault</strong></li><li><strong>Application insights</strong></li></ul><p>These components are essential for tracking experiments and managing your model securely.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*YqmzSbus3OamJ4zWnzzDIg.png" /></figure><h3>Step 3 — Launch Azure Machine Learning Studio</h3><p>Within the workspace, click <strong>“Launch Studio”</strong> to open Azure ML Studio — the UI where you’ll manage datasets, models, and experiments.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*d0ANzSb2rJa85ozawNcX0g.png" /></figure><h3>Step 4 — Start a New Automated ML Job</h3><ul><li>In the left-hand menu, go to <strong>“Automated ML”</strong></li><li>Click <strong>“+ New Automated ML job”</strong> to begin a new training experiment</li></ul><h3>Step 5 — Configure Basic Job Settings</h3><p>Fill in the basic settings such as:</p><ul><li>Experiment name</li><li>Compute target</li><li>Data storage</li></ul><p>Then, click <strong>“Next”</strong> to proceed.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*ZV-M7MTW4Mg_zdvxcghqrA.png" /></figure><h3>Step 6 — Select the Task Type</h3><p>For this use case, choose <strong>“Regression”</strong> as the task type. Since we want to <strong>predict a numerical value (number of rentals)</strong>, regression is the most appropriate option.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/594/1*wlS1szb_PcpLe4Z9MP8YAQ.png" /></figure><h3>Step 7 — Upload the Training Dataset</h3><p>Download the sample dataset from <a href="https://aka.ms/bike-rentals">this Microsoft link</a>, unzip it, and upload it to the workspace.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aMM5EIKXvK5W_yRBVUbqww.png" /></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*aGe27xMZa4K_43Re0CkPoQ.png" /></figure><h3>Step 8 — Choose the Dataset</h3><p>After uploading, select the dataset you just added and click <strong>“Next”</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*sgBXBPhVnskpuK5MsOZuBg.png" /></figure><h3>Step 9 — Configure Task Settings</h3><p>In this section:</p><ul><li>Select <strong>rentals</strong> as the <strong>target column</strong>. This is the value you want your model to predict.</li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*UmreSr88uxwQgokNYjHnHA.png" /></figure><h3>Step 10 — Set Experiment Limits</h3><p>Set resource limits for the AutoML job. Since this is a <strong>proof of concept (PoC)</strong>, you can use minimal settings:</p><ul><li>Limit the maximum number of iterations</li><li>Choose a small node size for the compute target</li></ul><p>This step controls how many model/parameter combinations Azure will try during the experiment.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*rLSyQvzmRqtmc6RZ9NqBEQ.png" /></figure><h3>Step 11 — Train/Test Data Split</h3><p>Decide how to split the data:</p><ul><li>Common splits are <strong>80/20</strong> or <strong>90/10</strong> for training and testing respectively.</li></ul><p>This helps the model learn while still reserving a portion of data to evaluate its performance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1014/1*X37V7khbNecRBvJcB-BerQ.png" /></figure><h3>Step 12 — Select Compute Target</h3><p>You can use a pre-existing compute target or let Azure create a default one for you. Choose one appropriate for your workload and budget.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*BZRopoHEqBRwhrLV9fFWlQ.png" /></figure><h3>Step 13 — Review and Submit</h3><p>Review all configurations, then click <strong>“Submit”</strong> to start the AutoML experiment. Azure will now begin training multiple models and rank them based on their performance.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*75aAYeQNBXqbbXWEfUsHzQ.png" /></figure><h3>Conclusion</h3><p>In this first part of the series, we’ve successfully:</p><ul><li>Set up the Azure ML environment</li><li>Created a workspace</li><li>Launched an AutoML experiment to predict bicycle rentals</li></ul><p>In the next part of the series, we’ll explore how to <strong>deploy the best model</strong>, create an <strong>endpoint</strong>, and test it with new data.</p><p>Stay tuned for ; Predicting Bicycle Rentals with Azure Machine Learning — Part 2</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1ee10199c10c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Postgre Identifier “table_name_abc” will be truncated to “table_name”]]></title>
            <link>https://serkansakinmaz.medium.com/postgre-identifier-table-name-abc-will-be-truncated-to-table-name-e9a1cf7421d7?source=rss-583460cf9605------2</link>
            <guid isPermaLink="false">https://medium.com/p/e9a1cf7421d7</guid>
            <category><![CDATA[identifier]]></category>
            <category><![CDATA[database-development]]></category>
            <category><![CDATA[posgresql]]></category>
            <dc:creator><![CDATA[Serkan SAKINMAZ]]></dc:creator>
            <pubDate>Mon, 04 Sep 2023 09:00:46 GMT</pubDate>
            <atom:updated>2023-09-04T09:00:46.421Z</atom:updated>
            <content:encoded><![CDATA[<p>In your application, when you automatically generate the table from code, the table name could be different from how you create it. The challenge is, that you can not see any errors. If you deep dive to database logs, you can see the info message like below</p><p>INFO:root: Database Message: NOTICE: identifier “table_name_abc” will be truncated to “table_name”</p><p>In PostgreSQL, the table identifier is 63 bytes. If your table names extend this value, Postgres will truncate the latest characters. You can see the sample INFO message at the below</p><p>There are two options to fix this issue</p><ul><li>Option 1 — Change the settings of PosgreSQL</li><li>Option 2 —Change the identifier to shorten the table name</li></ul><p>In the first option, you need to change the Postgre source code and compile it. I would not recommend it since it needs some administration effort.</p><p>As a second option, you can change the identifier which would be the easiest option</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*b0_acib4EQHuGQcN7TPzOA.png" /></figure><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e9a1cf7421d7" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to fetch JSON via URL and parse in Python]]></title>
            <link>https://serkansakinmaz.medium.com/how-to-fetch-json-via-url-and-parse-in-python-222bbee74dee?source=rss-583460cf9605------2</link>
            <guid isPermaLink="false">https://medium.com/p/222bbee74dee</guid>
            <category><![CDATA[json]]></category>
            <category><![CDATA[exchange-rate]]></category>
            <category><![CDATA[python]]></category>
            <category><![CDATA[url]]></category>
            <dc:creator><![CDATA[Serkan SAKINMAZ]]></dc:creator>
            <pubDate>Fri, 01 Sep 2023 08:27:50 GMT</pubDate>
            <atom:updated>2023-09-01T08:27:50.853Z</atom:updated>
            <content:encoded><![CDATA[<p>In this post, I am going to explain simple a code snippet that is very helpful for lots of applications. With Python, we are going to fetch JSON content from the URL and parse some fields</p><p>We will use Python as a programming language. In Python, to request the content of the URL, you can use the requests library. Let’s test the following code snippet</p><p>response = requests.get(‘<a href="https://api.exchangerate.host/latest&#39;">https://api.exchangerate.host/latest&#39;</a>)<br>data = response.json()<br>print(data)</p><p>This code snippet prints the following output</p><p>{<br> “motd”:{<br> “msg”:”If you or your company use this project or like what we doing, please consider backing us so we can continue maintaining and evolving this project.”,<br> “url”:”<a href="https://exchangerate.host/#/donate">https://exchangerate.host/#/donate</a>&quot;<br> },<br> “success”:true,<br> “base”:”EUR”,<br> “date”:”2023–09–01&quot;,<br> “rates”:{<br> “AED”:3.980607,<br> “AFN”:88.275247,<br> “ALL”:108.652558,<br> “AMD”:419.641174,</p><p>As a next step, what we want to do is to parse the data and show the USD exchange rate</p><p><strong>json</strong> library allows to show the specific content of the JSON</p><p>print(data[“rates”][“USD”])</p><p>It shows the USD/EUR rates: 1.084458</p><h3>Conclusion</h3><p>In this blog, we have implemented a sample application that fetches JSON via URL. As you see, it is a very simple approach with Python applications.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=222bbee74dee" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to tackle challenging situations as a solution architect?]]></title>
            <link>https://serkansakinmaz.medium.com/how-to-tackle-challenging-situations-as-a-solution-architect-adb62657b02e?source=rss-583460cf9605------2</link>
            <guid isPermaLink="false">https://medium.com/p/adb62657b02e</guid>
            <category><![CDATA[solution-architecture]]></category>
            <category><![CDATA[architecture]]></category>
            <category><![CDATA[solution-architect]]></category>
            <dc:creator><![CDATA[Serkan SAKINMAZ]]></dc:creator>
            <pubDate>Thu, 31 Aug 2023 12:42:39 GMT</pubDate>
            <atom:updated>2023-08-31T12:42:39.501Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/244/0*hXBxanTDD537Ce_v.png" /></figure><p>Software development is more than implementing the functional requirements based on the technical roadmap. Since development is teamwork, you need to interact with different project members.</p><p>In this blog, I will explain how to tackle challenging situations based on my experience. It doesn&#39;t mean my behaviour is the best for specific situations. I will mention different cases and how I behaved. You can review, and use it as an experience.</p><h3>How to find the ideal solution within multiple teams?</h3><p>While implementing software development projects, there are different stakeholders. The development team wants to implement the best architecture, product team wants to see the feature as soon as possible and would like to proceed with a cost-effective solution. In this case, how to find the best option.</p><p>Let’s elaborate on some details;</p><ul><li><strong>Roles in the team</strong>: Developer, Solution Architect, Product owner</li><li><strong>Situation: </strong>The development team wants a real-time processing solution. It takes more effort and increases time to production. However, the product owner thinks that offer complicates the solution, and thinks there is no necessity to follow this option.</li></ul><p>In this case, how solution architect find a solution to make the team happy?</p><p>First of all, you couldn’t make everyone happy. The best option is to be transparent with everyone and find the optimal option. Based on the analysis, the current system is implemented with batch processing, and the solution architect will do more analysis.</p><p>As a solution architect, you can list down all options and write pros/cons for all possible solutions. I won’t deep dive into all the technical details, we are just looking for the tactical methodology</p><p>Real-time processing <strong>advantages</strong></p><ul><li>Process data real-time</li><li>etc...</li></ul><p>Real-time processing <strong>disadvantages</strong></p><ul><li>7/24 Support and Monitoring</li><li>etc...</li></ul><p>Batch processing <strong>advantages</strong></p><ul><li>More stable</li><li>Needs less maintenance effort</li><li>etc...</li></ul><p>Batch processing <strong>disadvantages</strong></p><ul><li>Latency</li><li>asdasd</li></ul><p>You can see all the advantages and disadvantages. As a second step, what you can do is to put some comments for the listed items. For example;</p><ul><li>Latency does not matter for the business</li><li>The technical team doesn&#39;t have enough capacity to monitor real-time processes and fix immediately when the issue happens</li><li>etc…</li></ul><p>These are small tips for a decision about the final solution. It seems, that development can be done via batch processing option. If you clearly explain why the batch option is selected, it would be better to put the most important advantages and disadvantages. Hence, everyone aligns on the solution</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*vHWw9lheuovYATPg.jpg" /></figure><h3>How to behave in a failed project?</h3><p>In one investigation, it was said that %70 of software projects failed. In this subject, we won’t discuss the reason for failed projects. I will give one example of how to behave within a team if the project has failed. If the project is implemented to the customer, it would be a more challenging situation</p><p>Let’s elaborate on some details;</p><ul><li><strong>Roles in the team</strong>: Developer, Solution Architect, Project Manager, Customer (Business owner)</li><li><strong>Situation: </strong>The project is processing the documents in order to make them searchable in the search system. After 6 months of work, the application has been deployed. However, most of the document process has failed due to unexpected load to the application. The project has been handed over to the customer, but they are not happy about the application, They asked to fix the problem without any cost, and they are not happy about the work, and don’t want to sign any contract in the future.</li></ul><p>In this case, how do solution architects find a solution to convince the customer that they fix the problem and work together for the upcoming projects?</p><p>First of all, you need to do some internal meetings to raise the issue and pay attention to risks. Put all the problems on the table, but do not blame anyone. At this time, you have to tackle all issues as a team, real teams can be seen at this moment.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*7ZAxTSoBQzNoPg6H" /></figure><p>Once you align on the problem, it would be better to find the optimal solution that fixes the problem. You need to also get some support from the management team. If you need more capacity, feel free and ask for support from the management team. As I mentioned, this kind of issue usually happens.</p><p>As a second step, please be transparent to the customer. Explained the challenge and the reason for the problem. And you can commit, that you will fix the problem as soon as possible without any cost from the customer. As expected, the customer will make some noise, but they also want that the issue will be solved. You will show that you are solution-oriented.</p><p>Once the implementation has started to fix the problem, make a demo as much as possible to the customer. It might have fewer features, however, it is always better to move away from over-commitment. They might change some features.</p><h3>How to tackle long-term projects?</h3><p>In ideal scrum projects, your team implement the features and deploys them with 2–3 weeks iterations. When you start a transformation project, there will be more work you couldn’t foresee during the project. As a solution architect and project team, you might be given 6 months to complete the transformation. Most probably, it will fail in terms of timeline.</p><p>Let’s elaborate on some details;</p><ul><li><strong>Roles in the team</strong>: Developer, Solution Architect, Product Owner</li><li><strong>Situation: </strong>The company wants to move from the monolith database to a cloud-based database. It is not just a lift-shift approach, some of the components need to be transformed. The transformation team has 6 months to complete the project. After starting the project, the development team found more work to complete the transformation and it would take more than 1 year, even more.</li></ul><p>Once this issue happens, the project team ask to find a solution and approach to finish the project successfully.</p><p>First of all, you need to clearly explain that the project has more work than the initial estimation. It should be explained to everyone, that it is not a development or architecture issue, the problem is wrong estimation. Even if you make a good estimation, you need to be open to surprises.</p><p>In this case, I would recommend splitting the transformation project into smaller parts. It will accelerate your life.</p><p>Let’s assume you are planning to do the following transformation</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*qODkJ8T_7LNEO3HZ.png" /></figure><p>Instead of transforming everything at the same time, you can make transformation step by step. For example;</p><ul><li>Transformation Phase 1: Order Placement Transformation</li><li>Transformation Phase 2: Inventory Transformation</li><li>Transformation Phase 3: Analytics Transformation</li><li>Transformation Phase 4: Customer Communications Transformation</li></ul><p>Once you finish the first phase successfully, the team can start other phases. There might be still surprises, however, it shouldn’t be big like a full transformation.</p><h3>Conclusion</h3><p>As you see, there are different approaches to solving the challenges in the project. As a solution architect, you are not only responsible for architecture, you should also lead the project. There is no certain recommendation for each use case. There are some problems and solutions that happen in real project work.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=adb62657b02e" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to manage your data with Upstash]]></title>
            <link>https://serkansakinmaz.medium.com/how-to-manage-your-data-with-upstash-2203e905beb9?source=rss-583460cf9605------2</link>
            <guid isPermaLink="false">https://medium.com/p/2203e905beb9</guid>
            <category><![CDATA[serverless]]></category>
            <category><![CDATA[redis]]></category>
            <category><![CDATA[upstash]]></category>
            <category><![CDATA[kafka]]></category>
            <dc:creator><![CDATA[Serkan SAKINMAZ]]></dc:creator>
            <pubDate>Sun, 08 Jan 2023 09:12:45 GMT</pubDate>
            <atom:updated>2023-01-09T06:31:52.564Z</atom:updated>
            <content:encoded><![CDATA[<p><a href="https://redis.com/">Redis</a> is a distributed in-memory database that provides peak performance for applications. When you need a good caching solution, Redis comes to play. In terms of infrastructure management, you can manage Redis on-prem or on the cloud. In this blog, I am going to explain how to manage Redis data in a serverless manner with <a href="https://upstash.com/?utm_source=Serkan1">Upstash</a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*fmJ4W-EhJWO3ZZVar1zB9g.png" /></figure><p>Serverless is getting more popular in the cloud world. You don’t need to manage any server, and cloud providers introduce new serverless services that focus on specific domains and libraries. For example, Lambda is one of the popular serverless services that provides a computing solution.</p><p>There is also a need to manage Redis and Kafka data in a serverless manner. At that point, <a href="https://upstash.com/?utm_source=Serkan1">Upstash</a> provides a data platform that manages data for Kafka and Redis. It is serverless, hence you don’t need to provision any server. In addition to that, Upstash provides low latency to access data on these platforms.</p><p>As of now, Upstash supports Kafka, Redis and QStash, most probably they will support more solutions in the future.</p><h4>Advantages and features of <a href="https://upstash.com/?utm_source=Serkan1">Upstash</a></h4><ul><li>Supports pay-as-you-go pricing: When you use <a href="https://upstash.com/?utm_source=Serkan1">Upstash</a>, you don’t need to pay any money upfront or don’t need to sign a contract. The fee is calculated by how much you used the service. In addition to that, Upstash supports big enterprise companies with other offers</li><li>Easy to use: You can easily enrol on the service and start using it within minutes. We will implement a sample application to show the easiness.</li><li>Monitoring support: You can monitor different metrics like request count, cost, average storage etc .. Under the usage panel, you can see lots of details with your table</li><li>Availability and scalability: Upstash supports multi-zone replication to improve availability and scalability. It also meets with a disaster recovery strategy for critical applications.</li><li>Backup: With the backup feature, you can easily create an on-demand and scheduled backup. The backup can be restored to the latest backup time.</li><li>Security: It supports the different levels of data encryption to improve data security. Apart from that, Upstash supports some additional security features like IP Whitelisting, TLS etc</li><li>Reduce time to production: Since Upstash manages the infrastructure, you are going to focus on your product instead of focusing on server provisioning and management.</li><li>Web-based CLI: You can easily use Redis commands via CLI. The details are in this link: <a href="https://docs.upstash.com/redis/overall/rediscompatibility">https://docs.upstash.com/redis/overall/rediscompatibility</a></li><li>Web-based data browser: Upstash provides a web-based data browser, so you don’t need to install any client-based data viewer.</li><li>Documentation: It is very important for new beginners as well as developers who will use more complicated features. Upstash documentation includes mandatory details for users</li></ul><h4>Sample Application with Upstash</h4><p>Let’s implement a quick POC with Upstash. For beginners, Upstash provides a free tier plan, and you don’t need to pay any money.</p><p>Official link: <a href="https://upstash.com/">https://upstash.com/</a></p><p><strong>Step 1- </strong>Creating an account on Upstash</p><p><a href="https://upstash.com/?utm_source=Serkan1">Upstash</a> allows you to create an account with your email, password or other accounts like Amazon, GitHub and Google</p><figure><img alt="Sign-in to Upstash" src="https://cdn-images-1.medium.com/max/462/1*2AhjM9FLbSTfZqz-U2URKw.png" /></figure><p><strong>Step 2</strong>-Create a Redis database from the console. You can also import and migrate the database from the console</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1008/1*YXuHa2LyoPk-oGpIp0W5KA.png" /></figure><p><strong>Step 3</strong>-You are going to fill out the database name and regional information</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/716/1*9UqWkbVv3QjfaTVEqp3vyQ.png" /></figure><p><strong>Step 4</strong>-After creating the database, you will see the details to connect to the database</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5vzHpUz8usp0pfE29lxQNQ.png" /></figure><p><strong>Step 5</strong>-You can see the different types of code snippets. In the below panel, <a href="https://upstash.com/?utm_source=Serkan1">Upstash</a> also provides a REST API call</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*kbVuxfvCFfrVC8lUjjJWsQ.png" /></figure><p>Let’s implement a simple Java code to use the Redis database which is on <a href="https://upstash.com/?utm_source=Serkan1">Upstash</a>. As I mentioned, Upstash provides code snippets that support popular programming languages.</p><p>I will use Jedis as a Redis client library to connect Redis from Java, you can use the following maven snippet to import the Jedis library</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/db0303a24a292b8744523ee575ceaada/href">https://medium.com/media/db0303a24a292b8744523ee575ceaada/href</a></iframe><p>Create a simple Java code and paste the following which also exists in your <a href="https://upstash.com/?utm_source=Serkan1">Upstash</a> panel</p><iframe src="" width="0" height="0" frameborder="0" scrolling="no"><a href="https://medium.com/media/c30da2706bf4eeb297ee8f953f6e47ac/href">https://medium.com/media/c30da2706bf4eeb297ee8f953f6e47ac/href</a></iframe><p>Note: You need to replace your Redis URL and password.</p><p>The code connects to Redis which is in the <a href="https://upstash.com/?utm_source=Serkan1">Upstash</a> and adds a simple key/value pair like “UserPoint”</p><p>After running the code snippet, we will see the value that we put to Redis</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/836/1*Gy83fPBwn1S3SP-6TnwZXQ.png" /></figure><p>Congrats! You easily used Redis via Upstash within minutes. We have implemented a simple application with Redis.</p><p>For developers who want to use Kafka, it is still simple how we implemented it in Redis.</p><h3>Conclusion</h3><p>There are different architectural approaches to setting up your project infrastructure either in the on-premise or in the cloud. Cloud comes to play with different advantages like easiness, logging, monitoring, scalability and availability. If you don’t have any impediment to using the cloud, I think the cloud is a very good option. For using Redis and Kafka, <a href="https://upstash.com/?utm_source=Serkan1">Upstash</a> is one of the best solutions within the cloud providers with serverless support for that requirement.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=2203e905beb9" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>