Yapay zeka ajanlar\u0131, g\u00fcn\u00fcm\u00fcz\u00fcn h\u0131zla geli\u015fen teknoloji d\u00fcnyas\u0131nda giderek daha kritik bir rol oynamaktad\u0131r. Karma\u015f\u0131k g\u00f6revleri otomatize etme, karar verme s\u00fcre\u00e7lerini optimize etme ve insan-bilgisayar etkile\u015fimini d\u00f6n\u00fc\u015ft\u00fcrme potansiyeline sahip bu otonom sistemler, i\u015fletmeler i\u00e7in yeni ufuklar a\u00e7maktad\u0131r. Ancak, bu ajanlar\u0131 \u00f6l\u00e7eklenebilir, g\u00fcvenilir ve y\u00f6netilebilir bir \u015fekilde da\u011f\u0131tmak, \u00f6zellikle b\u00fcy\u00fck dil modellerinin (LLM) entegrasyonu s\u00f6z konusu oldu\u011funda \u00f6nemli zorluklar\u0131 beraberinde getirir. Bu makale, Kubernetes’in g\u00fcc\u00fcn\u00fc Kagent ile birle\u015ftirerek ve Amazon Bedrock’un zengin temel model yelpazesinden faydalanarak yapay zeka ajanlar\u0131n\u0131 nas\u0131l etkili bir \u015fekilde kurup y\u00f6netece\u011finizi detayl\u0131 bir \u015fekilde a\u00e7\u0131klayacakt\u0131r.<\/p>\n
Yapay zeka ajanlar\u0131, belirli hedeflere ula\u015fmak i\u00e7in ortamlar\u0131n\u0131 alg\u0131layan, bilgi i\u015fleyen ve eylemler ger\u00e7ekle\u015ftiren otonom yaz\u0131l\u0131m varl\u0131klar\u0131d\u0131r. Geleneksel yaz\u0131l\u0131mlar\u0131n aksine, ajanlar genellikle dinamik ortamlarda \u00e7al\u0131\u015f\u0131r, \u00f6\u011frenme yetene\u011fine sahiptir ve karma\u015f\u0131k g\u00f6revleri yerine getirmek i\u00e7in birden fazla ad\u0131m\u0131 koordine edebilirler. Bu \u00f6zellikleri onlar\u0131 m\u00fc\u015fteri hizmetlerinden finansal analize, sa\u011fl\u0131k hizmetlerinden lojisti\u011fe kadar geni\u015f bir yelpazede uygulamalar i\u00e7in ideal k\u0131lar.<\/p>\n
Bir yapay zeka ajan\u0131 genellikle \u015fu bile\u015fenlerden olu\u015fur: bir alg\u0131lay\u0131c\u0131 (sens\u00f6r) ile \u00e7evresinden veri toplar, bir muhakeme motoru (reasoning engine) ile bu verileri i\u015fler ve karar verir, bir haf\u0131za ile ge\u00e7mi\u015f deneyimlerini saklar ve bir eylemci (actuator) ile \u00e7evresinde de\u011fi\u015fiklikler yapar. B\u00fcy\u00fck Dil Modelleri (LLM’ler), bu muhakeme motorunun ve hatta haf\u0131za bile\u015feninin temelini olu\u015fturarak ajanlara do\u011fal dil anlama, \u00fcretme ve karma\u015f\u0131k problem \u00e7\u00f6zme yetenekleri kazand\u0131r\u0131r.<\/p>\n
Bir yapay zeka ajan\u0131 tek ba\u015f\u0131na etkileyici olabilir, ancak ger\u00e7ek d\u00fcnya uygulamalar\u0131nda genellikle y\u00fczlerce, hatta binlerce ajan\u0131n e\u015f zamanl\u0131 olarak \u00e7al\u0131\u015fmas\u0131 gerekir. \u00d6rne\u011fin, bir e-ticaret platformu binlerce m\u00fc\u015fteriye ayn\u0131 anda hizmet veren yapay zeka destekli sohbet botlar\u0131na ihtiya\u00e7 duyabilir. Bu senaryolarda \u00f6l\u00e7eklenebilirlik, sistemin artan talebi kar\u015f\u0131layabilmesi, performanstan \u00f6d\u00fcn vermeden daha fazla i\u015f y\u00fck\u00fcn\u00fc kald\u0131rabilmesi anlam\u0131na gelir. Y\u00fcksek trafik anlar\u0131nda bile h\u0131zl\u0131 yan\u0131t s\u00fcreleri sa\u011flamak ve maliyetleri optimize etmek i\u00e7in \u00f6l\u00e7eklenebilirlik kritik \u00f6neme sahiptir.<\/p>\n
Geleneksel sunucu tabanl\u0131 veya sanal makine tabanl\u0131 da\u011f\u0131t\u0131m y\u00f6ntemleri, yapay zeka ajanlar\u0131n\u0131n dinamik ve \u00f6l\u00e7eklenebilir gereksinimlerini kar\u015f\u0131lamakta zorlanabilir. Kaynak tahsisi, y\u00fck dengeleme, otomatik \u00f6l\u00e7eklendirme, ar\u0131za tolerans\u0131 ve s\u00fcr\u00fcm y\u00f6netimi gibi konular, manuel olarak y\u00f6netildi\u011finde karma\u015f\u0131k ve zaman al\u0131c\u0131 hale gelir. \u00d6zellikle LLM’ler gibi kaynak yo\u011fun modellerle \u00e7al\u0131\u015f\u0131rken, bu zorluklar daha da artar ve geli\u015ftiricilerin ana i\u015flerine odaklanmas\u0131n\u0131 engeller.<\/p>\n
Kubernetes, konteynerli i\u015f y\u00fcklerini ve hizmetlerini otomatikle\u015ftirmek, da\u011f\u0131tmak, \u00f6l\u00e7eklendirmek ve y\u00f6netmek i\u00e7in a\u00e7\u0131k kaynakl\u0131 bir platformdur. Yapay zeka ajanlar\u0131n\u0131n da\u011f\u0131t\u0131m\u0131ndaki geleneksel zorluklara g\u00fc\u00e7l\u00fc bir \u00e7\u00f6z\u00fcm sunar.<\/p>\n
Kagent, Kubernetes \u00fczerinde yapay zeka ajanlar\u0131n\u0131 y\u00f6netmek i\u00e7in \u00f6zel olarak tasarlanm\u0131\u015f bir \u00e7er\u00e7evedir. Geli\u015ftiricilerin, ajanlar\u0131n ya\u015fam d\u00f6ng\u00fcs\u00fcn\u00fc (olu\u015fturma, da\u011f\u0131tma, g\u00fcncelleme, silme) Kubernetes’in yerel mekanizmalar\u0131yla entegre bir \u015fekilde y\u00f6netmelerini sa\u011flar. Kagent, ajanlar\u0131n durumunu izler, kaynaklar\u0131n\u0131 optimize eder ve farkl\u0131 temel modellerle (\u00f6rne\u011fin Amazon Bedrock) sorunsuz bir \u015fekilde etkile\u015fim kurmalar\u0131na olanak tan\u0131r. Kagent, temelde bir Kubernetes Operat\u00f6r\u00fc olarak \u00e7al\u0131\u015f\u0131r ve \u00f6zel kaynak tan\u0131mlar\u0131 (Custom Resource Definitions – CRD) kullanarak ajanlar\u0131 bir Kubernetes objesi gibi ele almam\u0131z\u0131 sa\u011flar.<\/p>\n
Kagent, Kubernetes API’sini kullanarak ajanlar\u0131 y\u00f6netir. Bir geli\u015ftirici bir Kagent CRD’si tan\u0131mlad\u0131\u011f\u0131nda, Kagent Operat\u00f6r\u00fc bu tan\u0131m\u0131 alg\u0131lar ve buna g\u00f6re Kubernetes pod’lar\u0131, servisleri ve di\u011fer kaynaklar\u0131 olu\u015fturur. Bu entegrasyon, ajanlar\u0131n Kubernetes ekosisteminin t\u00fcm avantajlar\u0131ndan (otomatik \u00f6l\u00e7eklendirme, sa\u011fl\u0131k kontrolleri, a\u011f politikalar\u0131 vb.) faydalanmas\u0131n\u0131 sa\u011flar. Ayr\u0131ca, Kagent, ajanlar\u0131n d\u0131\u015f servislerle (bu durumda Amazon Bedrock) g\u00fcvenli bir \u015fekilde ileti\u015fim kurmas\u0131 i\u00e7in gerekli yap\u0131land\u0131rmalar\u0131 ve kimlik do\u011frulama mekanizmalar\u0131n\u0131 da y\u00f6netebilir.<\/p>\n
Amazon Bedrock, \u00f6nde gelen temel modelleri (Foundation Models – FM) tek bir API arac\u0131l\u0131\u011f\u0131yla eri\u015filebilir k\u0131lan, tam olarak y\u00f6netilen bir hizmettir. Bu, geli\u015ftiricilerin farkl\u0131 LLM’leri kolayca denemesine, da\u011f\u0131tmas\u0131na ve \u00f6l\u00e7eklendirmesine olanak tan\u0131r.<\/p>\n
Bedrock, Amazon’un kendi Titan modelleri de dahil olmak \u00fczere, Anthropic’in Claude’u, AI21 Labs’\u0131n Jurassic’i, Cohere’\u0131n Command ve Embed’i gibi \u00e7e\u015fitli pop\u00fcler temel modellere eri\u015fim sa\u011flar. Bu modeller, metin olu\u015fturma, \u00f6zetleme, \u00e7eviri, soru yan\u0131tlama ve hatta kod olu\u015fturma gibi geni\u015f bir yelpazede g\u00f6revleri yerine getirebilir. Bedrock, modelleri y\u00f6netmek i\u00e7in gerekli altyap\u0131y\u0131 soyutlayarak, geli\u015ftiricilerin model se\u00e7imi ve uygulama mant\u0131\u011f\u0131na odaklanmas\u0131n\u0131 sa\u011flar.<\/p>\n
Bedrock’un en b\u00fcy\u00fck avantajlar\u0131ndan biri, farkl\u0131 LLM’ler aras\u0131nda kolayca ge\u00e7i\u015f yapabilme yetene\u011fidir. Her modelin kendine \u00f6zg\u00fc g\u00fc\u00e7l\u00fc y\u00f6nleri ve kullan\u0131m durumlar\u0131 vard\u0131r. \u00d6rne\u011fin:<\/p>\n
Bu \u00e7e\u015fitlilik, Kagent ile entegre edildi\u011finde, ajanlar\u0131n\u0131z\u0131n belirli g\u00f6revler i\u00e7in en uygun modeli dinamik olarak se\u00e7mesine veya farkl\u0131 modelleri bir arada kullanmas\u0131na olanak tan\u0131r.<\/p>\n
Kagent ve Amazon Bedrock’un birle\u015fimi, Kubernetes \u00fczerinde g\u00fc\u00e7l\u00fc ve \u00f6l\u00e7eklenebilir yapay zeka ajanlar\u0131 olu\u015fturmak i\u00e7in sa\u011flam bir mimari sunar.<\/p>\n
Temel mimari \u015fu \u015fekildedir:<\/p>\n
Bu yap\u0131, ajanlar\u0131n Kubernetes’in \u00f6l\u00e7eklendirme ve y\u00f6netim yeteneklerinden faydalan\u0131rken, Bedrock’un sundu\u011fu geni\u015f LLM yelpazesine kolayca eri\u015fmesini sa\u011flar.<\/p>\n
Bir yapay zeka ajan\u0131, bir g\u00f6revi yerine getirirken \u015fu ad\u0131mlar\u0131 izleyebilir:<\/p>\n
anthropic.claude-v2<\/code>) ve prompt’u i\u00e7erir.<\/li>\n- Model Yan\u0131t\u0131:<\/strong> Bedrock, iste\u011fi i\u015fler ve se\u00e7ilen modelden gelen yan\u0131t\u0131 ajana geri d\u00f6nd\u00fcr\u00fcr.<\/li>\n
- Yan\u0131t \u0130\u015fleme ve Eylem:<\/strong> Ajan, modelden gelen yan\u0131t\u0131 i\u015fler (\u00f6rne\u011fin, \u00f6zetler, ayr\u0131\u015ft\u0131r\u0131r) ve buna g\u00f6re bir eylem ger\u00e7ekle\u015ftirir veya kullan\u0131c\u0131ya bir \u00e7\u0131kt\u0131 sunar.<\/li>\n<\/ol>\n
G\u00fcvenlik ve Kimlik Do\u011frulama Mekanizmalar\u0131<\/h4>\n
Ajanlar\u0131n Bedrock’a g\u00fcvenli bir \u015fekilde eri\u015fmesi kritik \u00f6neme sahiptir. AWS ortam\u0131nda bu genellikle IAM (Identity and Access Management) rolleri arac\u0131l\u0131\u011f\u0131yla sa\u011flan\u0131r. Kubernetes’te \u00e7al\u0131\u015fan pod’lar\u0131n AWS kaynaklar\u0131na eri\u015fimi i\u00e7in genellikle IRSA (IAM Roles for Service Accounts) kullan\u0131l\u0131r. Bu, her bir Kubernetes Service Account’\u0131na belirli bir IAM rol\u00fc atayarak, o Service Account’\u0131 kullanan pod’lar\u0131n ilgili AWS kaynaklar\u0131na (bu durumda Bedrock) eri\u015fim yetkisi kazanmas\u0131n\u0131 sa\u011flar. Bu sayede, hassas AWS kimlik bilgilerini pod’lar\u0131n i\u00e7ine g\u00f6mmeye gerek kalmaz.<\/p>\n
Kurulum Rehberi: Kagent ve Bedrock’u Kubernetes’e Da\u011f\u0131tma<\/h3>\n
Bu b\u00f6l\u00fcmde, Kagent’i Kubernetes k\u00fcmenize kurmak ve Amazon Bedrock ile entegre etmek i\u00e7in ad\u0131m ad\u0131m bir rehber sunulmaktad\u0131r.<\/p>\n
\u00d6n Ko\u015fullar ve Haz\u0131rl\u0131klar<\/h4>\n\n- Aktif bir AWS hesab\u0131.<\/li>\n
- AWS CLI’nin kurulu ve yap\u0131land\u0131r\u0131lm\u0131\u015f olmas\u0131.<\/li>\n
- Kubernetes k\u00fcmesi (\u00f6rne\u011fin, Amazon EKS).<\/li>\n
kubectl<\/code> ve helm<\/code> CLI ara\u00e7lar\u0131n\u0131n kurulu olmas\u0131.<\/li>\n- Bedrock hizmetine eri\u015fim yetkisi olan bir IAM rol\u00fc (veya kullan\u0131c\u0131). Bu rol\u00fcn
bedrock:InvokeModel<\/code> iznine sahip olmas\u0131 gerekmektedir.<\/li>\n<\/ul>\nKubernetes K\u00fcmesi Kurulumu (K\u0131sa)<\/h4>\n
E\u011fer bir Kubernetes k\u00fcmeniz yoksa, Amazon EKS (Elastic Kubernetes Service) kullanarak kolayca bir k\u00fcme olu\u015fturabilirsiniz. \u00d6rne\u011fin, eksctl<\/code> arac\u0131 ile:<\/p>\neksctl create cluster --name kagent-bedrock-cluster --region us-east-1 --node-type t3.medium --nodes 2<\/pre>\n<\/code><\/p>\n
Bu komut, us-east-1<\/code> b\u00f6lgesinde iki adet t3.medium<\/code> tipinde d\u00fc\u011f\u00fcme sahip bir EKS k\u00fcmesi olu\u015fturacakt\u0131r.<\/p>\nKagent'in Da\u011f\u0131t\u0131m\u0131<\/h4>\n
Kagent'i Kubernetes k\u00fcmenize Helm grafi\u011fi kullanarak da\u011f\u0131tabilirsiniz:<\/p>\n
\n- Kagent Helm deposunu ekleyin:<\/strong>\n
helm repo add kagent https:\/\/kagent.github.io\/helm-charts\nhelm repo update<\/pre>\n<\/code>\n <\/li>\n
- Kagent Operat\u00f6r\u00fcn\u00fc kurun:<\/strong>\n
helm install kagent kagent\/kagent-operator --namespace kagent-system --create-namespace<\/pre>\n<\/code><\/p>\n
Bu komut, kagent-system<\/code> ad\u0131nda bir namespace olu\u015fturacak ve Kagent Operat\u00f6r\u00fcn\u00fc bu namespace'e da\u011f\u0131tacakt\u0131r.<\/p>\n<\/li>\n- Kurulumu do\u011frulay\u0131n:<\/strong>\n
kubectl get pods -n kagent-system<\/pre>\n<\/code><\/p>\n
Kagent operat\u00f6r pod'unun \u00e7al\u0131\u015ft\u0131\u011f\u0131n\u0131 g\u00f6rmelisiniz.<\/p>\n<\/li>\n<\/ol>\n
Bedrock Eri\u015fimi ve API Yap\u0131land\u0131rmas\u0131 (IRSA ile)<\/h4>\n
Kubernetes pod'lar\u0131n\u0131z\u0131n Amazon Bedrock'a g\u00fcvenli bir \u015fekilde eri\u015fmesi i\u00e7in IRSA (IAM Roles for Service Accounts) yap\u0131land\u0131rmas\u0131 \u00f6nerilir:<\/p>\n
\n- OIDC Sa\u011flay\u0131c\u0131s\u0131n\u0131 etkinle\u015ftirin:<\/strong> EKS k\u00fcmeniz i\u00e7in bir OIDC (OpenID Connect) sa\u011flay\u0131c\u0131s\u0131 etkinle\u015ftirmeniz gerekir. E\u011fer zaten etkin de\u011filse,
eksctl<\/code> ile yapabilirsiniz:\neksctl utils associate-iam-oidc-provider --cluster kagent-bedrock-cluster --approve<\/pre>\n<\/code>\n <\/li>\n
- Bedrock eri\u015fimi i\u00e7in IAM rol\u00fc olu\u015fturun:<\/strong> Bu rol,
bedrock:InvokeModel<\/code> iznine sahip olmal\u0131d\u0131r.\n{\n \"Version\": \"2012-10-17\",\n \"Statement\": [\n {\n \"Effect\": \"Allow\",\n \"Action\": \"bedrock:InvokeModel\",\n \"Resource\": \"*\"\n }\n ]\n}<\/pre>\n<\/code><\/p>\n
Bu politikay\u0131 i\u00e7eren bir IAM rol\u00fc olu\u015fturun (\u00f6rne\u011fin, KagentBedrockAccessRole<\/code>).<\/p>\n<\/li>\n- Kubernetes Service Account ve IAM rol\u00fcn\u00fc ili\u015fkilendirin:<\/strong> Kagent'in kullanaca\u011f\u0131 Service Account'\u0131 olu\u015fturun ve bu Service Account'\u0131 yukar\u0131da olu\u015fturdu\u011funuz IAM rol\u00fc ile ili\u015fkilendirin.\n
eksctl create iamserviceaccount \\\n --cluster kagent-bedrock-cluster \\\n --namespace default \\\n --name kagent-bedrock-sa \\\n --attach-policy-arn arn:aws:iam::YOUR_AWS_ACCOUNT_ID:policy\/KagentBedrockAccessPolicy \\\n --approve \\\n --override-existing-serviceaccounts<\/pre>\n<\/code><\/p>\n
YOUR_AWS_ACCOUNT_ID<\/code> ve politika ARN'sini kendi bilgilerinizle de\u011fi\u015ftirin.<\/p>\n<\/li>\n<\/ol>\n\u00d6rnek Bir Ajan\u0131n \u00c7al\u0131\u015ft\u0131r\u0131lmas\u0131<\/h4>\n
\u015eimdi Kagent kullanarak Amazon Bedrock'u kullanan basit bir yapay zeka ajan\u0131 da\u011f\u0131tal\u0131m. A\u015fa\u011f\u0131daki YAML dosyas\u0131, Anthropic Claude modelini kullanan bir ajan\u0131 tan\u0131mlar:<\/p>\n
apiVersion: kagent.io\/v1alpha1\nkind: Agent\nmetadata:\n name: bedrock-claude-agent\n namespace: default\nspec:\n image: \"your-agent-image:latest\" # Kendi ajan uygulaman\u0131z\u0131n Docker imaj\u0131\n replicas: 1\n serviceAccountName: kagent-bedrock-sa # Bedrock eri\u015fimi olan Service Account\n env:\n - name: AWS_REGION\n value: \"us-east-1\" # Bedrock'un bulundu\u011fu b\u00f6lge\n - name: BEDROCK_MODEL_ID\n value: \"anthropic.claude-v2\" # Kullan\u0131lacak Bedrock modeli\n ports:\n - containerPort: 8080\n name: http\n # Ajan uygulaman\u0131z\u0131n Bedrock ile etkile\u015fim kodu 'your-agent-image' i\u00e7inde olmal\u0131.\n # \u00d6rnek Python kodu:\n # import boto3\n # import os\n #\n # bedrock_runtime = boto3.client(\n # service_name='bedrock-runtime',\n # region_name=os.environ.get('AWS_REGION', 'us-east-1')\n # )\n #\n # model_id = os.environ.get('BEDROCK_MODEL_ID', 'anthropic.claude-v2')\n # prompt = \"Merhaba, nas\u0131ls\u0131n?\"\n #\n # body = json.dumps({\n # \"prompt\": f\"\\n\\nHuman: {prompt}\\n\\nAssistant:\",\n # \"max_tokens_to_sample\": 300,\n # \"temperature\": 0.5,\n # \"top_p\": 0.9\n # })\n #\n # response = bedrock_runtime.invoke_model(\n # body=body,\n # modelId=model_id,\n # accept=\"application\/json\",\n # contentType=\"application\/json\"\n # )\n #\n # response_body = json.loads(response.get('body').read())\n # print(response_body.get('completion'))\n<\/pre>\n<\/code><\/p>\n
Bu YAML'i uygulad\u0131ktan sonra Kagent, belirtti\u011finiz Docker imaj\u0131n\u0131 kullanarak bir pod ba\u015flatacak ve bu pod, Bedrock'a eri\u015fim yetkisine sahip olacakt\u0131r. Kendi ajan uygulaman\u0131z\u0131n Docker imaj\u0131n\u0131 olu\u015fturman\u0131z ve i\u00e7inde Bedrock API'sini \u00e7a\u011f\u0131ran kodu bulundurman\u0131z gerekmektedir.<\/p>\n
kubectl apply -f bedrock-agent.yaml<\/pre>\n<\/code><\/p>\n
Pod'un durumunu kontrol edin:<\/p>\n
kubectl get pods -l kagent.io\/agent=bedrock-claude-agent<\/pre>\n<\/code><\/p>\n
En \u0130yi Uygulamalar ve \u0130pu\u00e7lar\u0131<\/h3>\n
Yapay zeka ajanlar\u0131n\u0131z\u0131 Kubernetes \u00fczerinde Kagent ve Bedrock ile da\u011f\u0131t\u0131rken a\u015fa\u011f\u0131daki en iyi uygulamalar\u0131 g\u00f6z \u00f6n\u00fcnde bulundurun:<\/p>\n
Performans Optimizasyonu<\/h4>\n\n- Do\u011fru Model Se\u00e7imi:<\/strong> G\u00f6revinize en uygun ve maliyet etkin Bedrock modelini se\u00e7in. Her modelin performans\u0131 ve maliyeti farkl\u0131d\u0131r.<\/li>\n
- Prompt M\u00fchendisli\u011fi:<\/strong> LLM'lerden en iyi yan\u0131tlar\u0131 almak i\u00e7in etkili prompt m\u00fchendisli\u011fi teknikleri kullan\u0131n.<\/li>\n
- Kaynak Tahsisi:<\/strong> Ajan pod'lar\u0131n\u0131z i\u00e7in do\u011fru CPU, bellek ve gerekirse GPU kaynaklar\u0131n\u0131 tahsis edin. A\u015f\u0131r\u0131 tahsis maliyetleri art\u0131r\u0131rken, yetersiz tahsis performans\u0131 d\u00fc\u015f\u00fcr\u00fcr.<\/li>\n
- \u00d6nbellekleme:<\/strong> Tekrarlayan veya s\u0131k kullan\u0131lan LLM \u00e7a\u011fr\u0131lar\u0131 i\u00e7in bir \u00f6nbellekleme katman\u0131 eklemeyi d\u00fc\u015f\u00fcn\u00fcn.<\/li>\n<\/ul>\n
\u0130zleme ve G\u00fcnl\u00fckleme<\/h4>\n\n- Prometheus\/Grafana:<\/strong> Kubernetes k\u00fcmenizin ve ajan pod'lar\u0131n\u0131z\u0131n metriklerini (CPU, bellek kullan\u0131m\u0131, a\u011f trafi\u011fi) izlemek i\u00e7in Prometheus ve Grafana kullan\u0131n.<\/li>\n
- Merkezi G\u00fcnl\u00fckleme:<\/strong> Ajanlar\u0131n\u0131z\u0131n \u00fcretti\u011fi g\u00fcnl\u00fckleri (loglar\u0131) merkezi bir sisteme (\u00f6rne\u011fin, Amazon CloudWatch Logs, Elasticsearch\/Kibana) g\u00f6nderin. Bu, sorun giderme ve performans analizi i\u00e7in kritik \u00f6neme sahiptir.<\/li>\n
- Bedrock Metrikleri:<\/strong> Amazon CloudWatch \u00fczerinden Bedrock API \u00e7a\u011fr\u0131lar\u0131n\u0131z\u0131n metriklerini (\u00e7a\u011fr\u0131 say\u0131s\u0131, gecikme s\u00fcresi, hata oranlar\u0131) izleyin.<\/li>\n<\/ul>\n
G\u00fcvenlik ve Eri\u015fim Y\u00f6netimi<\/h4>\n\n- En Az Ayr\u0131cal\u0131k Prensibi:<\/strong> IAM rollerinize ve Kubernetes Service Account'lar\u0131n\u0131za yaln\u0131zca ajanlar\u0131n ihtiya\u00e7 duydu\u011fu minimum yetkileri verin.<\/li>\n
- A\u011f Politikalar\u0131:<\/strong> Kubernetes a\u011f politikalar\u0131 (Network Policies) kullanarak ajan pod'lar\u0131n\u0131z\u0131n hangi di\u011fer pod'larla veya d\u0131\u015f servislerle ileti\u015fim kurabilece\u011fini k\u0131s\u0131tlay\u0131n.<\/li>\n
- Konteyner G\u00fcvenli\u011fi:<\/strong> Ajan imajlar\u0131n\u0131z\u0131 d\u00fczenli olarak g\u00fcvenlik a\u00e7\u0131klar\u0131 i\u00e7in taray\u0131n ve g\u00fcncel tutun. G\u00fcvenilir temel imajlar kullan\u0131n.<\/li>\n
- Hassas Veri Y\u00f6netimi:<\/strong> Hassas verileri do\u011frudan kodda veya yap\u0131land\u0131rma dosyalar\u0131nda saklamaktan ka\u00e7\u0131n\u0131n. Kubernetes Secrets veya AWS Secrets Manager gibi \u00e7\u00f6z\u00fcmler kullan\u0131n.<\/li>\n<\/ul>\n
Maliyet Y\u00f6netimi<\/h4>\n\n- Otomatik \u00d6l\u00e7eklendirme:<\/strong> Kubernetes'in yatay pod otomatik \u00f6l\u00e7ekleyicisini (HPA) kullanarak ajanlar\u0131n\u0131z\u0131n yaln\u0131zca ihtiya\u00e7 duyuldu\u011funda \u00f6l\u00e7eklenmesini sa\u011flay\u0131n.<\/li>\n
- Spot \u00d6rnekleri:<\/strong> Maliyetleri d\u00fc\u015f\u00fcrmek i\u00e7in kritik olmayan i\u015f y\u00fckleri i\u00e7in Kubernetes d\u00fc\u011f\u00fcmlerinde Spot \u00f6rneklerini kullanmay\u0131 d\u00fc\u015f\u00fcn\u00fcn.<\/li>\n
- Bedrock Maliyetlerini \u0130zleme:<\/strong> AWS Maliyet Y\u00f6netimi ara\u00e7lar\u0131n\u0131 kullanarak Bedrock kullan\u0131m\u0131n\u0131z\u0131 ve maliyetlerinizi d\u00fczenli olarak g\u00f6zden ge\u00e7irin.<\/li>\n<\/ul>\n
Sonu\u00e7<\/h3>\n
Yapay zeka ajanlar\u0131, i\u015fletmelerin dijital d\u00f6n\u00fc\u015f\u00fcm yolculu\u011funda \u00f6nemli bir kald\u0131ra\u00e7 sunmaktad\u0131r. Bu ajanlar\u0131 Kubernetes \u00fczerinde Kagent ile orkestra etmek ve Amazon Bedrock'un zengin temel model ekosisteminden faydalanmak, \u00f6l\u00e7eklenebilir, g\u00fcvenilir ve y\u00f6netilebilir bir altyap\u0131 olu\u015fturman\u0131n anahtar\u0131d\u0131r. Bu entegrasyon, geli\u015ftiricilerin altyap\u0131 karma\u015f\u0131kl\u0131\u011f\u0131yla u\u011fra\u015fmak yerine, ajanlar\u0131n\u0131n i\u015f mant\u0131\u011f\u0131na ve yenilik\u00e7i \u00f6zelliklerine odaklanmalar\u0131n\u0131 sa\u011flar. Kagent ve Amazon Bedrock'un birle\u015fimi, yapay zeka ajanlar\u0131n\u0131z\u0131 h\u0131zla \u00fcretime alman\u0131za ve gelecekteki b\u00fcy\u00fcme ihtiya\u00e7lar\u0131n\u0131za kolayca adapte olman\u0131za olanak tan\u0131yan g\u00fc\u00e7l\u00fc bir sinerji yarat\u0131r. Bu rehberdeki ad\u0131mlar\u0131 takip ederek, kendi yapay zeka ajanlar\u0131n\u0131z\u0131 Kubernetes'in g\u00fcc\u00fcyle donatabilir ve Amazon Bedrock'un ak\u0131ll\u0131 yetenekleriyle zenginle\u015ftirebilirsiniz.<\/p>\n
SSS (S\u0131k Sorulan Sorular)<\/h3>\n\n- Kagent nedir ve neden kullanmal\u0131y\u0131m?<\/dt>\n
- Kagent, Kubernetes \u00fczerinde yapay zeka ajanlar\u0131n\u0131n ya\u015fam d\u00f6ng\u00fcs\u00fcn\u00fc y\u00f6netmek i\u00e7in tasarlanm\u0131\u015f bir \u00e7er\u00e7evedir. Ajanlar\u0131n\u0131z\u0131 Kubernetes'in otomatik \u00f6l\u00e7eklendirme, y\u00fcksek eri\u015filebilirlik ve kaynak y\u00f6netimi gibi avantajlar\u0131ndan faydalanarak kolayca da\u011f\u0131tman\u0131z\u0131, g\u00fcncellemenizi ve izlemenizi sa\u011flar. Bu, geli\u015ftiricilerin altyap\u0131 yerine ajan mant\u0131\u011f\u0131na odaklanmas\u0131na olanak tan\u0131r.<\/dd>\n
- Amazon Bedrock hangi temel modelleri destekler?<\/dt>\n
- Amazon Bedrock, Amazon'un kendi Titan modelleri (Text, Embed) dahil olmak \u00fczere, Anthropic'in Claude'u, AI21 Labs'\u0131n Jurassic'i ve Cohere'\u0131n Command ve Embed modelleri gibi \u00e7e\u015fitli \u00f6nde gelen temel modelleri destekler. Bu modeller s\u00fcrekli g\u00fcncellenmekte ve yenileri eklenmektedir.<\/dd>\n
- Bu kurulumun maliyeti nedir?<\/dt>\n
- Maliyetler, kullan\u0131lan Kubernetes d\u00fc\u011f\u00fcmlerinin say\u0131s\u0131na ve tipine (EKS), Bedrock'ta \u00e7a\u011fr\u0131lan temel modellerin kullan\u0131m miktar\u0131na ve se\u00e7ilen AWS b\u00f6lgesine g\u00f6re de\u011fi\u015fir. Kubernetes altyap\u0131s\u0131 i\u00e7in EKS ve EC2 maliyetleri, Bedrock i\u00e7in ise \u00e7a\u011fr\u0131 ba\u015f\u0131na veya token ba\u015f\u0131na maliyetler s\u00f6z konusudur. En do\u011fru maliyet tahmini i\u00e7in AWS Fiyatland\u0131rma Hesaplay\u0131c\u0131s\u0131n\u0131 kullanman\u0131z ve Bedrock fiyatland\u0131rma sayfas\u0131n\u0131 incelemeniz \u00f6nerilir.<\/dd>\n
- G\u00fcvenlik nas\u0131l sa\u011flan\u0131r?<\/dt>\n
- G\u00fcvenlik, AWS IAM rolleri ve Kubernetes Service Account'lar\u0131 (IRSA ile) kullan\u0131larak sa\u011flan\u0131r. Bu sayede ajan pod'lar\u0131n\u0131z\u0131n Bedrock'a eri\u015fimi i\u00e7in en az ayr\u0131cal\u0131k prensibi uygulan\u0131r. Ayr\u0131ca, Kubernetes a\u011f politikalar\u0131, konteyner imaj g\u00fcvenli\u011fi ve hassas veri y\u00f6netimi i\u00e7in Kubernetes Secrets gibi mekanizmalarla genel g\u00fcvenlik art\u0131r\u0131labilir.<\/dd>\n
- Kagent'i kendi Kubernetes k\u00fcmemde \u00e7al\u0131\u015ft\u0131rabilir miyim?<\/dt>\n
- Evet, Kagent herhangi bir standart Kubernetes k\u00fcmesinde (\u00f6rne\u011fin, \u015firket i\u00e7i, Google Kubernetes Engine (GKE), Azure Kubernetes Service (AKS) veya Minikube gibi yerel bir k\u00fcme) \u00e7al\u0131\u015fabilir. \u00d6nemli olan, k\u00fcmenizin gerekli kaynaklara sahip olmas\u0131 ve AWS Bedrock ile entegrasyon i\u00e7in uygun kimlik do\u011frulama mekanizmalar\u0131n\u0131n yap\u0131land\u0131r\u0131lm\u0131\u015f olmas\u0131d\u0131r.<\/dd>\n<\/dl>\n","protected":false},"excerpt":{"rendered":"Yapay zeka ajanlar\u0131, g\u00fcn\u00fcm\u00fcz\u00fcn h\u0131zla geli\u015fen teknoloji d\u00fcnyas\u0131nda giderek daha kritik bir rol oynamaktad\u0131r. Karma\u015f\u0131k g\u00f6revleri otomatize etme, k…","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":[1491],"tags":[],"class_list":{"0":"post-38156","1":"post","2":"type-post","3":"status-publish","4":"format-standard","6":"category-kubernetes","7":"cs-entry","8":"cs-video-wrap"},"yoast_head":"\n
Kubernetes \u00dczerinde Yapay Zeka Ajanlar\u0131 Olu\u015fturma: Kagent ve Amazon Bedrock Kurulum Rehberi - 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\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/\" \/>\n<meta property=\"og:locale\" content=\"tr_TR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Kubernetes \u00dczerinde Yapay Zeka Ajanlar\u0131 Olu\u015fturma: Kagent ve Amazon Bedrock Kurulum Rehberi\" \/>\n<meta property=\"og:description\" content=\"Yapay zeka ajanlar\u0131, g\u00fcn\u00fcm\u00fcz\u00fcn h\u0131zla geli\u015fen teknoloji d\u00fcnyas\u0131nda giderek daha kritik bir rol oynamaktad\u0131r. Karma\u015f\u0131k g\u00f6revleri otomatize etme, k...\" \/>\n<meta property=\"og:url\" content=\"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/\" \/>\n<meta property=\"og:site_name\" content=\"Kodlar\u0131n Gizemli D\u00fcnyas\u0131\" \/>\n<meta property=\"article:published_time\" content=\"2026-01-24T13:30:40+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=\"16 dakika\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/\"},\"author\":{\"name\":\"Fatih Soysal\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\"},\"headline\":\"Kubernetes \u00dczerinde Yapay Zeka Ajanlar\u0131 Olu\u015fturma: Kagent ve Amazon Bedrock Kurulum Rehberi\",\"datePublished\":\"2026-01-24T13:30:40+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/\"},\"wordCount\":3065,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\"},\"articleSection\":[\"Kubernetes\"],\"inLanguage\":\"tr\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/#respond\"]}],\"copyrightYear\":\"2026\",\"copyrightHolder\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#organization\"}},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/\",\"url\":\"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/\",\"name\":\"Kubernetes \u00dczerinde Yapay Zeka Ajanlar\u0131 Olu\u015fturma: Kagent ve Amazon Bedrock Kurulum Rehberi - Kodlar\u0131n Gizemli D\u00fcnyas\u0131\",\"isPartOf\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#website\"},\"datePublished\":\"2026-01-24T13:30:40+00:00\",\"breadcrumb\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/#breadcrumb\"},\"inLanguage\":\"tr\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Anasayfa\",\"item\":\"https:\/\/fatihsoysal.com\/blog\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Kubernetes \u00dczerinde Yapay Zeka Ajanlar\u0131 Olu\u015fturma: Kagent ve Amazon Bedrock Kurulum Rehberi\"}]},{\"@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":"Kubernetes \u00dczerinde Yapay Zeka Ajanlar\u0131 Olu\u015fturma: Kagent ve Amazon Bedrock Kurulum Rehberi - 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\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/","og_locale":"tr_TR","og_type":"article","og_title":"Kubernetes \u00dczerinde Yapay Zeka Ajanlar\u0131 Olu\u015fturma: Kagent ve Amazon Bedrock Kurulum Rehberi","og_description":"Yapay zeka ajanlar\u0131, g\u00fcn\u00fcm\u00fcz\u00fcn h\u0131zla geli\u015fen teknoloji d\u00fcnyas\u0131nda giderek daha kritik bir rol oynamaktad\u0131r. Karma\u015f\u0131k g\u00f6revleri otomatize etme, k...","og_url":"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/","og_site_name":"Kodlar\u0131n Gizemli D\u00fcnyas\u0131","article_published_time":"2026-01-24T13:30:40+00:00","author":"Fatih Soysal","twitter_card":"summary_large_image","twitter_misc":{"Yazan:":"Fatih Soysal","Tahmini okuma s\u00fcresi":"16 dakika"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/#article","isPartOf":{"@id":"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/"},"author":{"name":"Fatih Soysal","@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1"},"headline":"Kubernetes \u00dczerinde Yapay Zeka Ajanlar\u0131 Olu\u015fturma: Kagent ve Amazon Bedrock Kurulum Rehberi","datePublished":"2026-01-24T13:30:40+00:00","mainEntityOfPage":{"@id":"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/"},"wordCount":3065,"commentCount":0,"publisher":{"@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1"},"articleSection":["Kubernetes"],"inLanguage":"tr","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/#respond"]}],"copyrightYear":"2026","copyrightHolder":{"@id":"https:\/\/fatihsoysal.com\/blog\/#organization"}},{"@type":"WebPage","@id":"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/","url":"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/","name":"Kubernetes \u00dczerinde Yapay Zeka Ajanlar\u0131 Olu\u015fturma: Kagent ve Amazon Bedrock Kurulum Rehberi - Kodlar\u0131n Gizemli D\u00fcnyas\u0131","isPartOf":{"@id":"https:\/\/fatihsoysal.com\/blog\/#website"},"datePublished":"2026-01-24T13:30:40+00:00","breadcrumb":{"@id":"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/#breadcrumb"},"inLanguage":"tr","potentialAction":[{"@type":"ReadAction","target":["https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/fatihsoysal.com\/blog\/kubernetes-uzerinde-yapay-zeka-ajanlari-olusturma-kagent-ve-amazon-bedrock-kurulum-rehberi\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Anasayfa","item":"https:\/\/fatihsoysal.com\/blog\/"},{"@type":"ListItem","position":2,"name":"Kubernetes \u00dczerinde Yapay Zeka Ajanlar\u0131 Olu\u015fturma: Kagent ve Amazon Bedrock Kurulum Rehberi"}]},{"@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\/38156","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=38156"}],"version-history":[{"count":0,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/posts\/38156\/revisions"}],"wp:attachment":[{"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/media?parent=38156"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/categories?post=38156"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/tags?post=38156"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}