{"id":36533,"date":"2025-12-19T13:00:36","date_gmt":"2025-12-19T10:00:36","guid":{"rendered":"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/"},"modified":"2025-12-19T13:00:36","modified_gmt":"2025-12-19T10:00:36","slug":"dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme","status":"publish","type":"post","link":"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/","title":{"rendered":"DEV Track Spotlight: Semantic Kernel ve Amazon Bedrock ile .NET Yapay Zeka Uygulamalar\u0131 Geli\u015ftirme"},"content":{"rendered":"<h2>DEV Track Spotlight: Semantic Kernel ve Amazon Bedrock ile .NET Yapay Zeka Uygulamalar\u0131 Geli\u015ftirme<\/h2>\n<p>G\u00fcn\u00fcm\u00fcz yaz\u0131l\u0131m d\u00fcnyas\u0131nda yapay zeka, uygulamalar\u0131n temel ta\u015flar\u0131ndan biri haline gelmi\u015f durumda. \u00d6zellikle b\u00fcy\u00fck dil modelleri (LLM&#8217;ler) sayesinde geli\u015ftiriciler, daha \u00f6nce hayal bile edilemeyen karma\u015f\u0131kl\u0131kta ve yetenekte uygulamalar olu\u015fturabiliyorlar. Bu makale, Microsoft&#8217;un Semantic Kernel SDK&#8217;s\u0131n\u0131 ve Amazon Web Services&#8217;\u0131n (AWS) g\u00fc\u00e7l\u00fc yapay zeka hizmeti Amazon Bedrock&#8217;u bir araya getirerek .NET ekosisteminde nas\u0131l \u00e7\u0131\u011f\u0131r a\u00e7an yapay zeka uygulamalar\u0131 geli\u015ftirebilece\u011finizi detayland\u0131racakt\u0131r. Bu iki teknolojinin sinerjisi, geli\u015ftiricilere hem esneklik hem de \u00f6l\u00e7eklenebilirlik sunarak yapay zeka projelerini bir \u00fcst seviyeye ta\u015f\u0131ma imkan\u0131 verir.<\/p>\n<h3>Semantic Kernel Nedir ve Neden \u00d6nemlidir?<\/h3>\n<p>Semantic Kernel (SK), Microsoft taraf\u0131ndan geli\u015ftirilen a\u00e7\u0131k kaynakl\u0131 bir SDK&#8217;d\u0131r. Temel amac\u0131, b\u00fcy\u00fck dil modellerini geleneksel programlama dilleriyle (C#, Python, Java gibi) entegre etmeyi kolayla\u015ft\u0131rmakt\u0131r. SK, LLM&#8217;leri sadece bir API \u00e7a\u011fr\u0131s\u0131 olmaktan \u00e7\u0131kar\u0131p, onlar\u0131 uygulamalar\u0131n\u0131z\u0131n do\u011fal bir par\u00e7as\u0131 haline getiren bir orkestrasyon katman\u0131 sunar. Bu sayede geli\u015ftiriciler, karma\u015f\u0131k yapay zeka i\u015f ak\u0131\u015flar\u0131n\u0131 daha d\u00fczenli ve y\u00f6netilebilir bir \u015fekilde olu\u015fturabilirler.<\/p>\n<h4>LLM Orkestrasyonunda Semantic Kernel&#8217;\u0131n Rol\u00fc<\/h4>\n<p>Semantic Kernel, LLM&#8217;lerin yeteneklerini, kendi yazd\u0131\u011f\u0131n\u0131z kodlarla (native code) ve harici servislerle (API&#8217;ler) birle\u015ftirerek ak\u0131ll\u0131 &#8220;agent&#8221;lar veya &#8220;plugin&#8221;ler olu\u015fturman\u0131za olanak tan\u0131r. Bu, LLM&#8217;lerin sadece metin \u00fcretmekle kalmay\u0131p, belirli g\u00f6revleri yerine getirmesini, ara\u00e7lar\u0131 kullanmas\u0131n\u0131 ve hatta planlama yapmas\u0131n\u0131 sa\u011flar. \u00d6rne\u011fin, bir LLM&#8217;e bir e-posta tasla\u011f\u0131 yazma g\u00f6revi verdi\u011finizde, SK bu g\u00f6revi bir &#8220;plugin&#8221; arac\u0131l\u0131\u011f\u0131yla e-posta g\u00f6nderme servisiyle entegre edebilir.<\/p>\n<h4>Temel Kavramlar: Kernel, Pluginler ve Bellek<\/h4>\n<ul>\n<li><strong>Kernel:<\/strong> Semantic Kernel&#8217;\u0131n kalbidir. LLM&#8217;lerle etkile\u015fimi y\u00f6netir, pluginleri \u00e7a\u011f\u0131r\u0131r ve bellek y\u00f6netimini sa\u011flar.<\/li>\n<li><strong>Pluginler (Skills):<\/strong> LLM&#8217;lerin belirli g\u00f6revleri yerine getirmesi i\u00e7in kulland\u0131\u011f\u0131 i\u015flevlerdir. Bunlar, LLM&#8217;in kendisinin bir par\u00e7as\u0131 olabilece\u011fi gibi (semantic functions) veya geleneksel kodunuzla yazd\u0131\u011f\u0131n\u0131z fonksiyonlar da (native functions) olabilir. \u00d6rne\u011fin, bir veritaban\u0131ndan bilgi \u00e7ekme veya bir API&#8217;ye istek g\u00f6nderme gibi g\u00f6revler i\u00e7in native pluginler yaz\u0131labilir.<\/li>\n<li><strong>Bellek (Memory):<\/strong> LLM&#8217;lerin \u00f6nceki etkile\u015fimleri veya harici kaynaklardan gelen bilgileri hat\u0131rlamas\u0131n\u0131 sa\u011flar. Bu, sohbet botlar\u0131 gibi durum bilgisi gerektiren uygulamalar i\u00e7in kritik \u00f6neme sahiptir. SK, vekt\u00f6r veritabanlar\u0131 gibi \u00e7e\u015fitli bellek \u00e7\u00f6z\u00fcmleriyle entegrasyon sunar.<\/li>\n<\/ul>\n<h3>Amazon Bedrock ile B\u00fcy\u00fck Dil Modellerine Eri\u015fim<\/h3>\n<p>Amazon Bedrock, AWS&#8217;nin tam olarak y\u00f6netilen bir hizmetidir ve lider yapay zeka \u015firketlerinin (Anthropic, AI21 Labs, Cohere, Stability AI) ve Amazon&#8217;un kendi geli\u015ftirdi\u011fi (Amazon Titan) y\u00fcksek performansl\u0131 temel modellerine (FM&#8217;ler) eri\u015fim sa\u011flar. Bedrock, bu modelleri g\u00fcvenli, \u00f6l\u00e7eklenebilir ve sunucusuz bir ortamda kullanman\u0131za olanak tan\u0131r, b\u00f6ylece altyap\u0131 y\u00f6netimiyle u\u011fra\u015fmak zorunda kalmazs\u0131n\u0131z.<\/p>\n<h4>Bedrock&#8217;un Sundu\u011fu Model \u00c7e\u015fitlili\u011fi ve Avantajlar\u0131<\/h4>\n<p>Amazon Bedrock, metin olu\u015fturma, \u00f6zetleme, \u00e7eviri, sohbet, g\u00f6r\u00fcnt\u00fc olu\u015fturma gibi \u00e7e\u015fitli g\u00f6revler i\u00e7in optimize edilmi\u015f geni\u015f bir model yelpazesi sunar. Bu modeller aras\u0131nda Anthropic Claude, AI21 Labs Jurassic-2, Cohere Command, Stability AI Stable Diffusion ve Amazon Titan modelleri bulunur. Bedrock&#8217;u kullanman\u0131n ba\u015fl\u0131ca avantajlar\u0131 \u015funlard\u0131r:<\/p>\n<ul>\n<li><strong>Model \u00c7e\u015fitlili\u011fi:<\/strong> Farkl\u0131 ihtiya\u00e7lar i\u00e7in farkl\u0131 modeller aras\u0131ndan se\u00e7im yapabilme.<\/li>\n<li><strong>Tam Y\u00f6netilen Hizmet:<\/strong> Altyap\u0131 y\u00f6netimi y\u00fck\u00fcn\u00fc ortadan kald\u0131r\u0131r.<\/li>\n<li><strong>G\u00fcvenlik ve Gizlilik:<\/strong> AWS&#8217;nin g\u00fc\u00e7l\u00fc g\u00fcvenlik \u00f6zellikleriyle verilerinizin g\u00fcvende kalmas\u0131n\u0131 sa\u011flar.<\/li>\n<li><strong>\u00d6l\u00e7eklenebilirlik:<\/strong> Uygulamalar\u0131n\u0131z\u0131n talepleri do\u011frultusunda otomatik olarak \u00f6l\u00e7eklenir.<\/li>\n<li><strong>\u00d6zelle\u015ftirme:<\/strong> Kendi verilerinizle modelleri ince ayar (fine-tuning) yapma imkan\u0131 sunar.<\/li>\n<\/ul>\n<h4>.NET Uygulamalar\u0131nda Amazon Bedrock Entegrasyonu<\/h4>\n<p>.NET geli\u015ftiricileri, AWS SDK for .NET arac\u0131l\u0131\u011f\u0131yla Amazon Bedrock&#8217;a kolayca eri\u015febilirler. Bu SDK, Bedrock API&#8217;lerini C# kodunuzdan \u00e7a\u011f\u0131rmak i\u00e7in gerekli t\u00fcm s\u0131n\u0131flar\u0131 ve y\u00f6ntemleri sa\u011flar. Bu entegrasyon, .NET uygulamalar\u0131n\u0131z\u0131n do\u011frudan Bedrock \u00fczerinde \u00e7al\u0131\u015fan LLM&#8217;lerin g\u00fcc\u00fcnden faydalanmas\u0131n\u0131 m\u00fcmk\u00fcn k\u0131lar.<\/p>\n<h3>Semantic Kernel ve Amazon Bedrock Entegrasyonu<\/h3>\n<p>Semantic Kernel ve Amazon Bedrock&#8217;un birle\u015fimi, .NET geli\u015ftiricileri i\u00e7in g\u00fc\u00e7l\u00fc bir yapay zeka geli\u015ftirme platformu sunar. Semantic Kernel, Bedrock&#8217;u bir &#8220;metin modeli&#8221; veya &#8220;g\u00f6mme modeli&#8221; sa\u011flay\u0131c\u0131s\u0131 olarak kullanarak, Bedrock&#8217;taki LLM&#8217;leri kendi orkestrasyon yetenekleriyle birle\u015ftirir.<\/p>\n<h4>Entegrasyon Ad\u0131mlar\u0131 ve Mimari<\/h4>\n<p>Bu entegrasyonu ger\u00e7ekle\u015ftirmek i\u00e7in a\u015fa\u011f\u0131daki ad\u0131mlar\u0131 izleyebilirsiniz:<\/p>\n<ol>\n<li><strong>AWS Hesab\u0131 ve Bedrock Eri\u015fimi:<\/strong> Bir AWS hesab\u0131na sahip olmal\u0131 ve Bedrock hizmetine eri\u015fim yetkileriniz oldu\u011fundan emin olmal\u0131s\u0131n\u0131z. Gerekirse, AWS IAM kullanarak uygun izinleri ayarlay\u0131n.<\/li>\n<li><strong>Gerekli NuGet Paketlerini Y\u00fckleme:<\/strong> .NET projenize Semantic Kernel ve Bedrock ba\u011flay\u0131c\u0131s\u0131 i\u00e7in gerekli NuGet paketlerini eklemelisiniz.<\/li>\n<li><strong>Kernel Yap\u0131land\u0131rmas\u0131:<\/strong> Semantic Kernel&#8217;\u0131 ba\u015flat\u0131rken, Amazon Bedrock&#8217;u bir metin modeli veya g\u00f6mme modeli olarak yap\u0131land\u0131rman\u0131z gerekir. Bu, AWS kimlik bilgilerinizi ve kullanmak istedi\u011finiz Bedrock modelinin ad\u0131n\u0131 belirtmeyi i\u00e7erir.<\/li>\n<\/ol>\n<p><strong>\u00d6rnek C# Yap\u0131land\u0131rmas\u0131:<\/strong><\/p>\n<pre><code class=\"language-csharp\">\nusing Microsoft.SemanticKernel;\nusing Microsoft.SemanticKernel.Connectors.AmazonBedrock;\nusing Microsoft.SemanticKernel.Connectors.AmazonBedrock.Core;\nusing Amazon;\nusing Amazon.Runtime;\n\n\/\/ AWS kimlik bilgilerini ve b\u00f6lgeyi ayarlay\u0131n\n\/\/ Bu bilgileri ortam de\u011fi\u015fkenlerinden, AWS CLI profilinden veya IAM rol\u00fcnden alabilirsiniz.\nvar awsCredentials = new BasicAWSCredentials(\"YOUR_ACCESS_KEY\", \"YOUR_SECRET_KEY\");\nvar awsRegion = RegionEndpoint.USEast1; \/\/ Kulland\u0131\u011f\u0131n\u0131z AWS b\u00f6lgesi\n\n\/\/ Kernel olu\u015fturucu ile Semantic Kernel'\u0131 yap\u0131land\u0131r\u0131n\nvar builder = Kernel.CreateBuilder();\n\n\/\/ Amazon Bedrock metin modelini ekleyin\nbuilder.AddAmazonBedrockTextGeneration(\n    modelId: \"anthropic.claude-v2\", \/\/ Kullanmak istedi\u011finiz Bedrock model ID'si\n    credentials: awsCredentials,\n    region: awsRegion\n);\n\n\/\/ Alternatif olarak, g\u00f6mme modeli eklemek i\u00e7in:\n\/\/ builder.AddAmazonBedrockTextEmbeddingGeneration(\n\/\/     modelId: \"amazon.titan-embed-text-v1\",\n\/\/     credentials: awsCredentials,\n\/\/     region: awsRegion\n\/\/ );\n\nvar kernel = builder.Build();\n\nConsole.WriteLine(\"Semantic Kernel, Amazon Bedrock ile ba\u015far\u0131yla yap\u0131land\u0131r\u0131ld\u0131.\");\n<\/pre>\n<p><\/code><\/p>\n<h3>Pratik Uygulama Senaryolar\u0131 ve Kod \u00d6rnekleri<\/h3>\n<p>Semantic Kernel ve Amazon Bedrock entegrasyonuyla geli\u015ftirilebilecek bir\u00e7ok pratik uygulama senaryosu bulunmaktad\u0131r. A\u015fa\u011f\u0131da basit bir metin \u00f6zetleme \"plugin\"inin nas\u0131l olu\u015fturulaca\u011f\u0131na dair bir \u00f6rnek bulunmaktad\u0131r.<\/p>\n<h4>Metin \u00d6zetleme Plugin'i Olu\u015fturma<\/h4>\n<p>Bu \u00f6rnekte, kullan\u0131c\u0131dan al\u0131nan uzun bir metni Amazon Bedrock'taki bir LLM arac\u0131l\u0131\u011f\u0131yla \u00f6zetleyen bir Semantic Kernel plugin'i olu\u015fturaca\u011f\u0131z.<\/p>\n<pre><code class=\"language-csharp\">\nusing Microsoft.SemanticKernel;\nusing Microsoft.SemanticKernel.Connectors.AmazonBedrock;\nusing Microsoft.SemanticKernel.Connectors.AmazonBedrock.Core;\nusing Amazon;\nusing Amazon.Runtime;\n\n\/\/ ... (Yukar\u0131daki Kernel yap\u0131land\u0131rma kodu ayn\u0131 kal\u0131r) ...\nvar awsCredentials = new BasicAWSCredentials(\"YOUR_ACCESS_KEY\", \"YOUR_SECRET_KEY\");\nvar awsRegion = RegionEndpoint.USEast1;\n\nvar builder = Kernel.CreateBuilder();\nbuilder.AddAmazonBedrockTextGeneration(\n    modelId: \"anthropic.claude-v2\", \n    credentials: awsCredentials,\n    region: awsRegion\n);\nvar kernel = builder.Build();\n\n\/\/ 1. Semantic Function (Plugin) Olu\u015fturma\n\/\/ Bu, LLM'e ne yapmas\u0131 gerekti\u011fini s\u00f6yleyen bir prompt'tur.\nvar summarizeFunction = kernel.CreateFunctionFromPrompt(\n    \"A\u015fa\u011f\u0131daki metni 3-4 c\u00fcmleyle \u00f6zetle:\\n\\n{input}\",\n    functionName: \"SummarizeText\",\n    description: \"Verilen metni \u00f6zetler.\"\n);\n\n\/\/ 2. Fonksiyonu \u00c7al\u0131\u015ft\u0131rma\nConsole.WriteLine(\"L\u00fctfen \u00f6zetlemek istedi\u011finiz metni girin (bitirmek i\u00e7in bo\u015f sat\u0131r b\u0131rak\u0131n):\");\nstring fullText = \"\";\nstring line;\nwhile (!string.IsNullOrEmpty(line = Console.ReadLine()))\n{\n    fullText += line + \"\\n\";\n}\n\nif (!string.IsNullOrWhiteSpace(fullText))\n{\n    var result = await kernel.InvokeAsync(summarizeFunction, new() { [\"input\"] = fullText });\n    Console.WriteLine(\"\\n\u00d6zet:\");\n    Console.WriteLine(result.GetValue<string>());\n}\nelse\n{\n    Console.WriteLine(\"\u00d6zetlenecek metin girilmedi.\");\n}\n<\/pre>\n<p><\/code><\/p>\n<p>Bu \u00f6rnek, Semantic Kernel'\u0131n prompt tabanl\u0131 fonksiyonlar\u0131 nas\u0131l kolayca olu\u015fturup \u00e7al\u0131\u015ft\u0131rabildi\u011fini g\u00f6stermektedir. <code>CreateFunctionFromPrompt<\/code> metodu, do\u011fal dil komutlar\u0131n\u0131 bir LLM'in anlayabilece\u011fi bir i\u015flevselli\u011fe d\u00f6n\u00fc\u015ft\u00fcr\u00fcr. <code>InvokeAsync<\/code> metodu ise bu i\u015flevi belirlenen LLM (bu durumda Amazon Bedrock'taki Claude modeli) \u00fczerinde \u00e7al\u0131\u015ft\u0131r\u0131r.<\/p>\n<h3>Geli\u015ftirme S\u00fcrecinde Dikkat Edilmesi Gerekenler<\/h3>\n<p>Yapay zeka uygulamalar\u0131 geli\u015ftirirken baz\u0131 \u00f6nemli fakt\u00f6rleri g\u00f6z \u00f6n\u00fcnde bulundurmak, projenizin ba\u015far\u0131s\u0131 i\u00e7in kritik \u00f6neme sahiptir.<\/p>\n<h4>Prompt M\u00fchendisli\u011fi ve Model Se\u00e7imi<\/h4>\n<p>LLM'lerden en iyi \u00e7\u0131kt\u0131y\u0131 almak i\u00e7in etkili prompt m\u00fchendisli\u011fi \u015fartt\u0131r. Prompt'lar\u0131n\u0131z\u0131 a\u00e7\u0131k, net ve hedefe y\u00f6nelik bir \u015fekilde tasarlamak, istenen sonu\u00e7lar\u0131 elde etmenizi sa\u011flar. Ayr\u0131ca, kullanaca\u011f\u0131n\u0131z modelin (\u00f6rne\u011fin, Claude, Jurassic-2, Titan) uygulaman\u0131z\u0131n gereksinimlerine en uygun olan\u0131 oldu\u011fundan emin olun. Her modelin kendine \u00f6zg\u00fc g\u00fc\u00e7l\u00fc ve zay\u0131f y\u00f6nleri vard\u0131r.<\/p>\n<h4>Maliyet Optimizasyonu<\/h4>\n<p>LLM kullan\u0131m\u0131n\u0131n maliyetleri token say\u0131s\u0131na g\u00f6re de\u011fi\u015fir. Prompt'lar\u0131n\u0131z\u0131 ve \u00e7\u0131kt\u0131lar\u0131n\u0131z\u0131 m\u00fcmk\u00fcn oldu\u011funca k\u0131sa ve \u00f6z tutarak maliyetleri optimize edebilirsiniz. Bedrock'un fiyatland\u0131rma modelini anlamak ve gereksiz API \u00e7a\u011fr\u0131lar\u0131ndan ka\u00e7\u0131nmak \u00f6nemlidir. Gerekirse, \u00f6nbellekleme (caching) stratejileri uygulayarak tekrar eden sorgular i\u00e7in LLM \u00e7a\u011fr\u0131lar\u0131n\u0131 azaltabilirsiniz.<\/p>\n<h4>G\u00fcvenlik ve Veri Gizlili\u011fi<\/h4>\n<p>\u00d6zellikle hassas verilerle \u00e7al\u0131\u015f\u0131rken g\u00fcvenlik ve veri gizlili\u011fi en \u00fcst d\u00fczeyde tutulmal\u0131d\u0131r. Amazon Bedrock, verilerinizin AWS altyap\u0131s\u0131 i\u00e7inde kalmas\u0131n\u0131 ve \u015fifrelenmesini sa\u011flar. IAM rolleri ve politikalar\u0131 ile Bedrock eri\u015fimini s\u0131k\u0131 bir \u015fekilde kontrol edin. Uygulaman\u0131z\u0131n g\u00fcvenlik a\u00e7\u0131klar\u0131na kar\u015f\u0131 d\u00fczenli olarak denetlenmesi de \u00f6nemlidir.<\/p>\n<h4>Performans ve \u00d6l\u00e7eklenebilirlik<\/h4>\n<p>Amazon Bedrock, y\u00fcksek performans ve \u00f6l\u00e7eklenebilirlik i\u00e7in tasarlanm\u0131\u015ft\u0131r. Ancak, uygulaman\u0131z\u0131n mimarisi de bu fakt\u00f6rleri desteklemelidir. Asenkron programlama desenlerini kullanarak API \u00e7a\u011fr\u0131lar\u0131n\u0131 engellemeyin. Yo\u011fun y\u00fck alt\u0131nda sisteminizin nas\u0131l davranaca\u011f\u0131n\u0131 test etmek i\u00e7in performans testleri yap\u0131n.<\/p>\n<p><strong>Tablo: Geli\u015ftirme S\u00fcrecinde Anahtar Fakt\u00f6rler<\/strong><\/p>\n<table>\n<thead>\n<tr>\n<th>Fakt\u00f6r<\/th>\n<th>A\u00e7\u0131klama<\/th>\n<th>En \u0130yi Uygulama<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Prompt M\u00fchendisli\u011fi<\/td>\n<td>LLM'den do\u011fru ve ilgili \u00e7\u0131kt\u0131y\u0131 alma sanat\u0131.<\/td>\n<td>A\u00e7\u0131k, spesifik ve ba\u011flam sa\u011flayan prompt'lar kullan\u0131n.<\/td>\n<\/tr>\n<tr>\n<td>Model Se\u00e7imi<\/td>\n<td>Uygulama gereksinimlerine uygun LLM'i se\u00e7me.<\/td>\n<td>Farkl\u0131 modelleri test edin, maliyet ve performans dengesini g\u00f6zetin.<\/td>\n<\/tr>\n<tr>\n<td>Maliyet Optimizasyonu<\/td>\n<td>LLM kullan\u0131m maliyetlerini d\u00fc\u015f\u00fcrme.<\/td>\n<td>Token kullan\u0131m\u0131n\u0131 minimize edin, \u00f6nbellekleme stratejileri uygulay\u0131n.<\/td>\n<\/tr>\n<tr>\n<td>G\u00fcvenlik<\/td>\n<td>Veri gizlili\u011fi ve sistem b\u00fct\u00fcnl\u00fc\u011f\u00fcn\u00fc koruma.<\/td>\n<td>IAM politikalar\u0131n\u0131 s\u0131k\u0131ca uygulay\u0131n, veri \u015fifrelemesini kullan\u0131n.<\/td>\n<\/tr>\n<tr>\n<td>Performans<\/td>\n<td>Uygulaman\u0131n h\u0131zl\u0131 ve duyarl\u0131 \u00e7al\u0131\u015fmas\u0131n\u0131 sa\u011flama.<\/td>\n<td>Asenkron \u00e7a\u011fr\u0131lar, uygun kaynak tahsisi, y\u00fck testi.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h3>Sonu\u00e7<\/h3>\n<p>Semantic Kernel ve Amazon Bedrock'un birle\u015fimi, .NET geli\u015ftiricileri i\u00e7in yapay zeka uygulamalar\u0131 geli\u015ftirme s\u00fcrecini basitle\u015ftiren ve g\u00fc\u00e7lendiren ola\u011fan\u00fcst\u00fc bir ara\u00e7 seti sunar. Semantic Kernel'\u0131n esnek orkestrasyon yetenekleri ile Amazon Bedrock'un geni\u015f ve y\u00f6netilen temel model portf\u00f6y\u00fc, geli\u015ftiricilere karma\u015f\u0131k yapay zeka \u00e7\u00f6z\u00fcmlerini h\u0131zl\u0131 ve verimli bir \u015fekilde hayata ge\u00e7irme imkan\u0131 tan\u0131r. \u0130ster ak\u0131ll\u0131 sohbet botlar\u0131, ister otomatik i\u00e7erik olu\u015fturma ara\u00e7lar\u0131, isterse karma\u015f\u0131k veri analizi sistemleri olsun, bu entegrasyon sayesinde .NET ekosisteminde s\u0131n\u0131rlar\u0131 zorlayan yapay zeka uygulamalar\u0131 in\u015fa edebilirsiniz. Bu teknolojileri ke\u015ffetmeye ve projelerinize entegre etmeye ba\u015flamak, sizi yapay zeka devriminin \u00f6n saflar\u0131na ta\u015f\u0131yacakt\u0131r.<\/p>\n<h3>S\u0131k Sorulan Sorular (SSS)<\/h3>\n<h4>H3>1. Semantic Kernel ile Amazon Bedrock aras\u0131ndaki temel fark nedir?<\/h4>\n<p>Semantic Kernel, LLM'leri geleneksel kod ve di\u011fer ara\u00e7larla entegre etmek i\u00e7in bir orkestrasyon katman\u0131 ve SDK'd\u0131r. Amazon Bedrock ise, \u00e7e\u015fitli temel modellere (LLM'ler dahil) API arac\u0131l\u0131\u011f\u0131yla eri\u015fim sa\u011flayan, tam y\u00f6netilen bir AWS hizmetidir. K\u0131sacas\u0131, Bedrock size LLM'leri sa\u011flar, Semantic Kernel ise bu LLM'leri uygulaman\u0131zda nas\u0131l kullanaca\u011f\u0131n\u0131z\u0131 y\u00f6netmenize yard\u0131mc\u0131 olur.<\/p>\n<h4>2. Hangi .NET s\u00fcr\u00fcmleri Semantic Kernel ve Amazon Bedrock entegrasyonunu destekler?<\/h4>\n<p>Semantic Kernel ve AWS SDK for .NET genellikle .NET 6 ve \u00fczeri s\u00fcr\u00fcmleri destekler. En g\u00fcncel ve g\u00fcvenli deneyim i\u00e7in .NET 8 kullan\u0131lmas\u0131 \u00f6nerilir.<\/p>\n<h4>3. Amazon Bedrock kullan\u0131m\u0131n\u0131n maliyetleri nas\u0131l y\u00f6netilir?<\/h4>\n<p>Maliyetler genellikle kullan\u0131lan modelin t\u00fcr\u00fcne ve i\u015flenen token say\u0131s\u0131na (giri\u015f ve \u00e7\u0131k\u0131\u015f) g\u00f6re belirlenir. Maliyetleri optimize etmek i\u00e7in prompt'lar\u0131n\u0131z\u0131 k\u0131sa tutun, gereksiz LLM \u00e7a\u011fr\u0131lar\u0131ndan ka\u00e7\u0131n\u0131n ve uygulaman\u0131zda \u00f6nbellekleme stratejileri kullanmay\u0131 d\u00fc\u015f\u00fcn\u00fcn. AWS maliyet y\u00f6netim ara\u00e7lar\u0131n\u0131 da d\u00fczenli olarak kontrol etmelisiniz.<\/p>\n<h4>4. Kendi \u00f6zel temel modellerimi Amazon Bedrock'a entegre edebilir miyim?<\/h4>\n<p>Evet, Amazon Bedrock, kendi \u00f6zel modellerinizi veya mevcut modelleri kendi verilerinizle ince ayar (fine-tuning) yaparak \u00f6zelle\u015ftirmenize olanak tan\u0131r. Bu sayede, uygulaman\u0131z\u0131n \u00f6zel ihtiya\u00e7lar\u0131na daha uygun \u00e7\u0131kt\u0131lar \u00fcretebilirsiniz.<\/p>\n<h4>5. Prompt m\u00fchendisli\u011fi neden bu kadar \u00f6nemli?<\/h4>\n<p>Prompt m\u00fchendisli\u011fi, LLM'lerden istenen, do\u011fru ve y\u00fcksek kaliteli \u00e7\u0131kt\u0131lar elde etmek i\u00e7in kritik \u00f6neme sahiptir. K\u00f6t\u00fc tasarlanm\u0131\u015f bir prompt, alakas\u0131z veya yanl\u0131\u015f sonu\u00e7lar do\u011furabilir. \u0130yi bir prompt, LLM'in g\u00f6revini, ba\u011flam\u0131 ve beklenen \u00e7\u0131kt\u0131 format\u0131n\u0131 a\u00e7\u0131k\u00e7a belirtir, b\u00f6ylece modelin performans\u0131n\u0131 maksimize eder.<\/p>\n","protected":false},"excerpt":{"rendered":"Semantic Kernel (SK), Microsoft taraf\u0131ndan geli\u015ftirilen a\u00e7\u0131k kaynakl\u0131 bir SDK&#8217;d\u0131r. Temel amac\u0131, b\u00fcy\u00fck dil modellerini geleneksel programlama dilleri&#8230;","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"csco_page_header_type":"","csco_page_load_nextpost":"","csco_page_subscribe_form":"","csco_page_contact_form":"","footnotes":""},"categories":[1340],"tags":[],"class_list":{"0":"post-36533","1":"post","2":"type-post","3":"status-publish","4":"format-standard","6":"category-net","7":"cs-entry","8":"cs-video-wrap"},"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v20.5 (Yoast SEO v25.3.1) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>DEV Track Spotlight: Semantic Kernel ve Amazon Bedrock ile .NET Yapay Zeka Uygulamalar\u0131 Geli\u015ftirme - Kodlar\u0131n Gizemli D\u00fcnyas\u0131<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/\" \/>\n<meta property=\"og:locale\" content=\"tr_TR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"DEV Track Spotlight: Semantic Kernel ve Amazon Bedrock ile .NET Yapay Zeka Uygulamalar\u0131 Geli\u015ftirme\" \/>\n<meta property=\"og:description\" content=\"Semantic Kernel (SK), Microsoft taraf\u0131ndan geli\u015ftirilen a\u00e7\u0131k kaynakl\u0131 bir SDK&#039;d\u0131r. Temel amac\u0131, b\u00fcy\u00fck dil modellerini geleneksel programlama dilleri...\" \/>\n<meta property=\"og:url\" content=\"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/\" \/>\n<meta property=\"og:site_name\" content=\"Kodlar\u0131n Gizemli D\u00fcnyas\u0131\" \/>\n<meta property=\"article:published_time\" content=\"2025-12-19T10:00:36+00:00\" \/>\n<meta name=\"author\" content=\"Fatih Soysal\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Yazan:\" \/>\n\t<meta name=\"twitter:data1\" content=\"Fatih Soysal\" \/>\n\t<meta name=\"twitter:label2\" content=\"Tahmini okuma s\u00fcresi\" \/>\n\t<meta name=\"twitter:data2\" content=\"11 dakika\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/\"},\"author\":{\"name\":\"Fatih Soysal\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\"},\"headline\":\"DEV Track Spotlight: Semantic Kernel ve Amazon Bedrock ile .NET Yapay Zeka Uygulamalar\u0131 Geli\u015ftirme\",\"datePublished\":\"2025-12-19T10:00:36+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/\"},\"wordCount\":1858,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\"},\"articleSection\":[\".NET\"],\"inLanguage\":\"tr\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/#respond\"]}],\"copyrightYear\":\"2025\",\"copyrightHolder\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#organization\"}},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/\",\"url\":\"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/\",\"name\":\"DEV Track Spotlight: Semantic Kernel ve Amazon Bedrock ile .NET Yapay Zeka Uygulamalar\u0131 Geli\u015ftirme - Kodlar\u0131n Gizemli D\u00fcnyas\u0131\",\"isPartOf\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#website\"},\"datePublished\":\"2025-12-19T10:00:36+00:00\",\"breadcrumb\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/#breadcrumb\"},\"inLanguage\":\"tr\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Anasayfa\",\"item\":\"https:\/\/fatihsoysal.com\/blog\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"DEV Track Spotlight: Semantic Kernel ve Amazon Bedrock ile .NET Yapay Zeka Uygulamalar\u0131 Geli\u015ftirme\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#website\",\"url\":\"https:\/\/fatihsoysal.com\/blog\/\",\"name\":\"Fatihsoysal.com\",\"description\":\"Blog - Yaz\u0131l\u0131m D\u00fcnyas\u0131 Tecr\u00fcbelerim\",\"publisher\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/fatihsoysal.com\/blog\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"tr\"},{\"@type\":[\"Person\",\"Organization\"],\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\",\"name\":\"Fatih Soysal\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"tr\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/fatihsoysal.com\/blog\/wp-content\/uploads\/2024\/04\/cropped-replicate-prediction-3kgg1hgjn5rgp0cf0p5tr0jw7w-1.png\",\"contentUrl\":\"https:\/\/fatihsoysal.com\/blog\/wp-content\/uploads\/2024\/04\/cropped-replicate-prediction-3kgg1hgjn5rgp0cf0p5tr0jw7w-1.png\",\"width\":512,\"height\":512,\"caption\":\"Fatih Soysal\"},\"logo\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/image\/\"},\"description\":\"Kullan\u0131m ve kodlama m\u00fckemmeliyetini odak alan uygulamalar olu\u015fturma deneyimine sahip, profesyonel olarak 15+ y\u0131l \u00fczeri deneyime sahip bir yaz\u0131l\u0131m m\u00fchendisi.\",\"url\":\"https:\/\/fatihsoysal.com\/blog\/author\/fatihsoysal\/\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"DEV Track Spotlight: Semantic Kernel ve Amazon Bedrock ile .NET Yapay Zeka Uygulamalar\u0131 Geli\u015ftirme - Kodlar\u0131n Gizemli D\u00fcnyas\u0131","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/","og_locale":"tr_TR","og_type":"article","og_title":"DEV Track Spotlight: Semantic Kernel ve Amazon Bedrock ile .NET Yapay Zeka Uygulamalar\u0131 Geli\u015ftirme","og_description":"Semantic Kernel (SK), Microsoft taraf\u0131ndan geli\u015ftirilen a\u00e7\u0131k kaynakl\u0131 bir SDK'd\u0131r. Temel amac\u0131, b\u00fcy\u00fck dil modellerini geleneksel programlama dilleri...","og_url":"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/","og_site_name":"Kodlar\u0131n Gizemli D\u00fcnyas\u0131","article_published_time":"2025-12-19T10:00:36+00:00","author":"Fatih Soysal","twitter_card":"summary_large_image","twitter_misc":{"Yazan:":"Fatih Soysal","Tahmini okuma s\u00fcresi":"11 dakika"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/#article","isPartOf":{"@id":"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/"},"author":{"name":"Fatih Soysal","@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1"},"headline":"DEV Track Spotlight: Semantic Kernel ve Amazon Bedrock ile .NET Yapay Zeka Uygulamalar\u0131 Geli\u015ftirme","datePublished":"2025-12-19T10:00:36+00:00","mainEntityOfPage":{"@id":"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/"},"wordCount":1858,"commentCount":0,"publisher":{"@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1"},"articleSection":[".NET"],"inLanguage":"tr","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/#respond"]}],"copyrightYear":"2025","copyrightHolder":{"@id":"https:\/\/fatihsoysal.com\/blog\/#organization"}},{"@type":"WebPage","@id":"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/","url":"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/","name":"DEV Track Spotlight: Semantic Kernel ve Amazon Bedrock ile .NET Yapay Zeka Uygulamalar\u0131 Geli\u015ftirme - Kodlar\u0131n Gizemli D\u00fcnyas\u0131","isPartOf":{"@id":"https:\/\/fatihsoysal.com\/blog\/#website"},"datePublished":"2025-12-19T10:00:36+00:00","breadcrumb":{"@id":"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/#breadcrumb"},"inLanguage":"tr","potentialAction":[{"@type":"ReadAction","target":["https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/fatihsoysal.com\/blog\/dev-track-spotlight-semantic-kernel-ve-amazon-bedrock-ile-net-yapay-zeka-uygulamalari-gelistirme\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Anasayfa","item":"https:\/\/fatihsoysal.com\/blog\/"},{"@type":"ListItem","position":2,"name":"DEV Track Spotlight: Semantic Kernel ve Amazon Bedrock ile .NET Yapay Zeka Uygulamalar\u0131 Geli\u015ftirme"}]},{"@type":"WebSite","@id":"https:\/\/fatihsoysal.com\/blog\/#website","url":"https:\/\/fatihsoysal.com\/blog\/","name":"Fatihsoysal.com","description":"Blog - Yaz\u0131l\u0131m D\u00fcnyas\u0131 Tecr\u00fcbelerim","publisher":{"@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/fatihsoysal.com\/blog\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"tr"},{"@type":["Person","Organization"],"@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1","name":"Fatih Soysal","image":{"@type":"ImageObject","inLanguage":"tr","@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/image\/","url":"https:\/\/fatihsoysal.com\/blog\/wp-content\/uploads\/2024\/04\/cropped-replicate-prediction-3kgg1hgjn5rgp0cf0p5tr0jw7w-1.png","contentUrl":"https:\/\/fatihsoysal.com\/blog\/wp-content\/uploads\/2024\/04\/cropped-replicate-prediction-3kgg1hgjn5rgp0cf0p5tr0jw7w-1.png","width":512,"height":512,"caption":"Fatih Soysal"},"logo":{"@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/image\/"},"description":"Kullan\u0131m ve kodlama m\u00fckemmeliyetini odak alan uygulamalar olu\u015fturma deneyimine sahip, profesyonel olarak 15+ y\u0131l \u00fczeri deneyime sahip bir yaz\u0131l\u0131m m\u00fchendisi.","url":"https:\/\/fatihsoysal.com\/blog\/author\/fatihsoysal\/"}]}},"yoast_meta":{"yoast_wpseo_title":"","yoast_wpseo_metadesc":"","yoast_wpseo_canonical":""},"amp_enabled":true,"_links":{"self":[{"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/posts\/36533","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/comments?post=36533"}],"version-history":[{"count":0,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/posts\/36533\/revisions"}],"wp:attachment":[{"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/media?parent=36533"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/categories?post=36533"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/tags?post=36533"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}