diff --git a/.github/workflows/kubectl_ai_build.yaml b/.github/workflows/kubectl_ai_build.yaml
new file mode 100644
index 000000000..0a695e4e0
--- /dev/null
+++ b/.github/workflows/kubectl_ai_build.yaml
@@ -0,0 +1,23 @@
+name: Kubeclt-ai Image Build on Push
+
+on:
+  workflow_dispatch:
+    inputs:
+      tag:
+        description: "The tag to apply to this build. Defaults to 'latest'."
+        default: "latest"
+        required: false
+  push:
+    paths:
+      - gen3-integration-tests/gen3_ci/kubectl_ai/**
+
+jobs:
+  AmazonLinuxBase2023:
+    name: Build Helm Cleanup Image
+    uses: uc-cdis/.github/.github/workflows/image_build_push.yaml@master
+    with:
+      DOCKERFILE_LOCATION: "./gen3-integration-tests/gen3_ci/kubectl_ai/Dockerfile"
+      OVERRIDE_REPO_NAME: "kubectl_ai"
+      OVERRIDE_TAG_NAME: "$(echo ${{ github.event_name == 'workflow_dispatch' && github.event.inputs.tag || format('{0}', github.head_ref || github.ref_name) }} | tr / _)"
+      USE_QUAY_ONLY: true
+    secrets: inherit
diff --git a/gen3-integration-tests/gen3_ci/kubectl_ai/Chart.yaml b/gen3-integration-tests/gen3_ci/kubectl_ai/Chart.yaml
new file mode 100644
index 000000000..b9b0459e5
--- /dev/null
+++ b/gen3-integration-tests/gen3_ci/kubectl_ai/Chart.yaml
@@ -0,0 +1,24 @@
+apiVersion: v2
+name: kubectl-ai
+description: A Helm chart for kubeclt-ai
+
+# A chart can be either an 'application' or a 'library' chart.
+#
+# Application charts are a collection of templates that can be packaged into versioned archives
+# to be deployed.
+#
+# Library charts provide useful utilities or functions for the chart developer. They're included as
+# a dependency of application charts to inject those utilities and functions into the rendering
+# pipeline. Library charts do not define any templates and therefore cannot be deployed.
+type: application
+
+# This is the chart version. This version number should be incremented each time you make changes
+# to the chart and its templates, including the app version.
+# Versions are expected to follow Semantic Versioning (https://semver.org/)
+version: 0.1.0
+
+# This is the version number of the application being deployed. This version number should be
+# incremented each time you make changes to the application. Versions are not expected to
+# follow Semantic Versioning. They should reflect the version the application is using.
+# It is recommended to use it with quotes.
+appVersion: "master"
diff --git a/gen3-integration-tests/gen3_ci/kubectl_ai/Dockerfile b/gen3-integration-tests/gen3_ci/kubectl_ai/Dockerfile
new file mode 100644
index 000000000..ba208f4b8
--- /dev/null
+++ b/gen3-integration-tests/gen3_ci/kubectl_ai/Dockerfile
@@ -0,0 +1,25 @@
+ARG AZLINUX_BASE_VERSION=master
+
+FROM quay.io/cdis/amazonlinux-base:${AZLINUX_BASE_VERSION} AS base
+
+# Install curl
+RUN dnf install -y --allowerasing \
+    curl \
+    findutils \
+    zstd \
+    tar \
+    gzip
+
+# Create kubectl directory
+WORKDIR /kubectl
+
+# Install kubectl-ai
+RUN curl -fsSL https://raw.githubusercontent.com/GoogleCloudPlatform/kubectl-ai/main/install.sh -o install.sh \
+    && sed -i 's/sudo //g' install.sh \
+    && bash install.sh
+
+# Install Kubectl
+RUN curl -LO https://dl.k8s.io/release/v1.33.0/bin/linux/amd64/kubectl && \
+    install -o root -g root -m 0755 kubectl /usr/local/bin/kubectl
+
+CMD ["sleep", "infinity"]
diff --git a/gen3-integration-tests/gen3_ci/kubectl_ai/templates/_helpers.tpl b/gen3-integration-tests/gen3_ci/kubectl_ai/templates/_helpers.tpl
new file mode 100644
index 000000000..a152c3c20
--- /dev/null
+++ b/gen3-integration-tests/gen3_ci/kubectl_ai/templates/_helpers.tpl
@@ -0,0 +1,6 @@
+{{/*
+Return the fully qualified app name
+*/}}
+{{- define "kubectl-ai.fullname" -}}
+{{- printf "%s" .Release.Name -}}
+{{- end }}
diff --git a/gen3-integration-tests/gen3_ci/kubectl_ai/templates/deployment.yaml b/gen3-integration-tests/gen3_ci/kubectl_ai/templates/deployment.yaml
new file mode 100644
index 000000000..ba6ac3e99
--- /dev/null
+++ b/gen3-integration-tests/gen3_ci/kubectl_ai/templates/deployment.yaml
@@ -0,0 +1,28 @@
+apiVersion: apps/v1
+kind: Deployment
+metadata:
+  name: kubectl-ai
+spec:
+  replicas: 1
+  selector:
+    matchLabels:
+      app: kubectl-ai
+  template:
+    metadata:
+      labels:
+        app: kubectl-ai
+    spec:
+      serviceAccountName: kubectl-access-kubectl-ai
+      containers:
+        - name: kubectl-ai
+          image: "{{ .Values.image.repository }}:{{ .Values.image.tag }}"
+          imagePullPolicy: {{ .Values.image.pullPolicy }}
+          resources:
+            {{- toYaml .Values.resources | nindent 12 }}
+          ports:
+            - containerPort: 80
+          env:
+            - name: OLLAMA_KEEP_ALIVE
+              value: "-1"
+            - name: KUBECTL_AI_TIMEOUT
+              value: "300"
diff --git a/gen3-integration-tests/gen3_ci/kubectl_ai/templates/kubectl-ai-pdb.yaml b/gen3-integration-tests/gen3_ci/kubectl_ai/templates/kubectl-ai-pdb.yaml
new file mode 100644
index 000000000..5f66810e8
--- /dev/null
+++ b/gen3-integration-tests/gen3_ci/kubectl_ai/templates/kubectl-ai-pdb.yaml
@@ -0,0 +1,11 @@
+{{- if .Values.pdb }}
+apiVersion: policy/v1
+kind: PodDisruptionBudget
+metadata:
+  name: kubectl-ai-pdb
+spec:
+  maxUnavailable: 0
+  selector:
+    matchLabels:
+       app.kubernetes.io/name: "kubectl-ai"
+{{- end }}
diff --git a/gen3-integration-tests/gen3_ci/kubectl_ai/templates/service-account.yaml b/gen3-integration-tests/gen3_ci/kubectl_ai/templates/service-account.yaml
new file mode 100644
index 000000000..6cee55888
--- /dev/null
+++ b/gen3-integration-tests/gen3_ci/kubectl_ai/templates/service-account.yaml
@@ -0,0 +1,37 @@
+apiVersion: v1
+kind: ServiceAccount
+metadata:
+  name: kubectl-access-kubectl-ai
+  namespace: {{ .Release.Namespace }}
+  annotations:
+    eks.amazonaws.com/role-arn: {{ .Values.serviceAccount.iamRoleArn }}
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: Role
+metadata:
+  name: kubectl-access-role-kubectl-ai
+  namespace: {{ .Release.Namespace }}
+rules:
+  - apiGroups: [""]
+    resources: ["pods", "pods/exec", "pods/log", "configmaps", "namespaces", "labels"]
+    verbs: ["get", "list", "create", "patch"]
+  - apiGroups: ["batch"]
+    resources: ["cronjobs", "jobs"]
+    verbs: ["get", "list", "create", "delete", "watch", "patch"]
+  - apiGroups: ["apps"]
+    resources: [deployments, deployments/scale]
+    verbs: ["get", "list", "create", "delete", "watch", "patch"]
+---
+apiVersion: rbac.authorization.k8s.io/v1
+kind: RoleBinding
+metadata:
+  name: kubectl-access-binding-kubectl-ai
+  namespace: {{ .Release.Namespace }}
+roleRef:
+  apiGroup: rbac.authorization.k8s.io
+  kind: Role
+  name: kubectl-access-role-kubectl-ai
+subjects:
+  - kind: ServiceAccount
+    name: kubectl-access-kubectl-ai
+    namespace: {{ .Release.Namespace }}
diff --git a/gen3-integration-tests/gen3_ci/kubectl_ai/templates/service.yaml b/gen3-integration-tests/gen3_ci/kubectl_ai/templates/service.yaml
new file mode 100644
index 000000000..19420ece8
--- /dev/null
+++ b/gen3-integration-tests/gen3_ci/kubectl_ai/templates/service.yaml
@@ -0,0 +1,11 @@
+apiVersion: v1
+kind: Service
+metadata:
+  name: kubectl-ai
+spec:
+  type: {{ .Values.service.type }}
+  ports:
+    - port: {{ .Values.service.port }}
+      targetPort: {{ .Values.service.port }}
+  selector:
+    app: kubectl-ai
diff --git a/gen3-integration-tests/gen3_ci/kubectl_ai/values.yaml b/gen3-integration-tests/gen3_ci/kubectl_ai/values.yaml
new file mode 100644
index 000000000..6879da69e
--- /dev/null
+++ b/gen3-integration-tests/gen3_ci/kubectl_ai/values.yaml
@@ -0,0 +1,23 @@
+image:
+  repository: quay.io/cdis/kubectl_ai
+  tag: chore_add_kubectl_ai
+  pullPolicy: Always
+
+pdb: true
+
+resources:
+  requests:
+    cpu: 2
+    memory: "2Gi"
+  limits:
+    cpu: 4
+    memory: "4Gi"
+
+service:
+  type: ClusterIP
+  port: 80
+
+# Service account for accessing S3 bedrock access
+serviceAccount:
+  name: s3-bedrock-access
+  iamRoleArn: arn:aws:iam::707767160287:role/kubectl-ai-irsa
diff --git a/gen3-integration-tests/gen3_ci/ollama/Dockerfile b/gen3-integration-tests/gen3_ci/ollama/Dockerfile
index c595b47e5..107d86d38 100644
--- a/gen3-integration-tests/gen3_ci/ollama/Dockerfile
+++ b/gen3-integration-tests/gen3_ci/ollama/Dockerfile
@@ -7,7 +7,8 @@ RUN dnf install -y --allowerasing \
     curl \
     findutils \
     zstd \
-    tar
+    tar \
+    gzip
 
 # Create model directory
 WORKDIR /models
@@ -22,6 +23,7 @@ RUN bash -c "\
     ollama pull gemma4:e4b \
 "
 
+
 # List ollama model
 RUN bash -c "\
     ollama serve & \
diff --git a/gen3-integration-tests/gen3_ci/ollama/templates/deployment.yaml b/gen3-integration-tests/gen3_ci/ollama/templates/deployment.yaml
index 2a5b472fe..d043d34e0 100644
--- a/gen3-integration-tests/gen3_ci/ollama/templates/deployment.yaml
+++ b/gen3-integration-tests/gen3_ci/ollama/templates/deployment.yaml
@@ -12,6 +12,7 @@ spec:
       labels:
         app: ollama
     spec:
+      serviceAccountName: kubectl-access-ollama
       containers:
         - name: ollama
           image: "{{ .Values.image.repository }}:{{ .Values.image.tag }}"
@@ -25,3 +26,7 @@ spec:
               value: "debug"
             - name: OLLAMA_HOST
               value: "0.0.0.0:11434"
+            - name: OLLAMA_KEEP_ALIVE
+              value: "-1"
+            - name: KUBECTL_AI_TIMEOUT
+              value: "300"
diff --git a/gen3-integration-tests/gen3_ci/ollama/values.yaml b/gen3-integration-tests/gen3_ci/ollama/values.yaml
index 6e316036d..68c18a370 100644
--- a/gen3-integration-tests/gen3_ci/ollama/values.yaml
+++ b/gen3-integration-tests/gen3_ci/ollama/values.yaml
@@ -1,17 +1,17 @@
 image:
   repository: quay.io/cdis/ollama
   tag: master
-  pullPolicy: IfNotPresent
+  pullPolicy: Always
 
 pdb: true
 
 resources:
   requests:
     cpu: 4
-    memory: "10Gi"
+    memory: "8Gi"
   limits:
     cpu: 6
-    memory: "12Gi"
+    memory: "10Gi"
 
 service:
   type: ClusterIP
diff --git a/gen3-integration-tests/gen3_ci/scripts/analyze_failed_tests_using_ai_agent.py b/gen3-integration-tests/gen3_ci/scripts/analyze_failed_tests_using_ai_agent.py
index a806abe67..d98ee32ac 100644
--- a/gen3-integration-tests/gen3_ci/scripts/analyze_failed_tests_using_ai_agent.py
+++ b/gen3-integration-tests/gen3_ci/scripts/analyze_failed_tests_using_ai_agent.py
@@ -13,12 +13,12 @@
 load_dotenv()
 
 
-def setup_ollama_helm_chart():
+def setup_helm_chart(service):
     cmd = [
         "helm",
         "install",
-        "ollama",
-        "gen3_ci/ollama",
+        service,
+        f"gen3_ci/{service}",
         "-n",
         os.getenv("NAMESPACE"),
     ]
@@ -31,7 +31,7 @@ def setup_ollama_helm_chart():
     )
     if not helm_install_result.returncode == 0:
         raise Exception(
-            f"Unable to install ollama. Error: {helm_install_result.stderr.strip()}"
+            f"Unable to install {service}. Error: {helm_install_result.stderr.strip()}"
         )
 
     cmd = [
@@ -40,21 +40,21 @@ def setup_ollama_helm_chart():
         "--for=condition=ready",
         "pod",
         "-l",
-        "app=ollama",
+        f"app={service}",
         "--timeout=10m",
         "-n",
         os.getenv("NAMESPACE"),
     ]
 
-    ollama_pod_ready_result = subprocess.run(
+    service_pod_ready_result = subprocess.run(
         cmd,
         stdout=subprocess.PIPE,
         stderr=subprocess.PIPE,
         text=True,
     )
-    if not ollama_pod_ready_result.returncode == 0:
+    if not service_pod_ready_result.returncode == 0:
         raise Exception(
-            f"Ollama pod hasn't started yet. Error: {ollama_pod_ready_result.stderr.strip()}"
+            f"{service} pod hasn't started yet. Error: {service_pod_ready_result.stderr.strip()}"
         )
 
 
@@ -69,11 +69,11 @@ def wait_for_port(host="localhost", port=11434, timeout=60):
     raise TimeoutError("Port-forward did not become ready")
 
 
-def setup_port_forwarding():
+def setup_port_forwarding(service):
     cmd = [
         "kubectl",
         "port-forward",
-        "svc/ollama",
+        f"svc/{service}",
         "11434:11434",
         "-n",
         os.getenv("NAMESPACE"),
@@ -88,11 +88,11 @@ def setup_port_forwarding():
     return process
 
 
-def uninstall_ollama_helm_chart():
+def uninstall_helm_chart(service):
     cmd = [
         "helm",
         "uninstall",
-        "ollama",
+        service,
         "-n",
         os.getenv("NAMESPACE"),
     ]
@@ -105,7 +105,7 @@ def uninstall_ollama_helm_chart():
     )
     if not helm_install_result.returncode == 0:
         raise Exception(
-            f"Unable to uninstall ollama. Error: {helm_install_result.stderr.strip()}"
+            f"Unable to uninstall {service}. Error: {helm_install_result.stderr.strip()}"
         )
 
 
@@ -149,7 +149,7 @@ def analyze_env_setup_failure() -> str:
         {"role": "system", "content": debug_prompt},
         {"role": "user", "content": "analyse the errors from this logfile"},
     ]
-    payload = {"model": "gemma4:e4b", "messages": messages, "temperature": 0}
+    payload = {"model": "qwen3.5:2b", "messages": messages, "temperature": 0}
     headers = {"Content-Type": "application/json"}
     url = "http://localhost:11434/v1/chat/completions"
     response = requests.post(url, json=payload, headers=headers)
@@ -158,6 +158,61 @@ def analyze_env_setup_failure() -> str:
     return response.content
 
 
+def analyze_env_setup_failure_using_kubectl_ai() -> str:
+    cmd = [
+        "kubectl",
+        "-n",
+        os.getenv("NAMESPACE"),
+        "get",
+        "pods",
+        "-l",
+        "app=kubectl-ai",
+        "-o",
+        "jsonpath='{.items[0].metadata.name}'",
+    ]
+    kubeclt_ai_pod_result = subprocess.run(
+        cmd,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.PIPE,
+        text=True,
+        timeout=600,
+    )
+    if not kubeclt_ai_pod_result.returncode == 0:
+        raise Exception(
+            f"Failed to get kubectl-ai pod. Error: {kubeclt_ai_pod_result.stderr.strip()}"
+        )
+    kubeclt_ai_pod_name = kubeclt_ai_pod_result.stdout.strip().replace("'", "")
+    kubectl_ai_cmd = [
+        "kubectl",
+        "-n",
+        os.getenv("NAMESPACE"),
+        "exec",
+        kubeclt_ai_pod_name,
+        "--",
+        "kubectl-ai",
+        "--llm-provider",
+        "ollama",
+        "--model",
+        "qwen3.5:2b",
+        "--skip-permissions",
+        f'Check if any pods are not healthy on {os.getenv("NAMESPACE")} namespace and anaylyze the logs',
+    ]
+    logger.info(kubectl_ai_cmd)
+    kubectl_ai_result = subprocess.run(
+        kubectl_ai_cmd,
+        stdout=subprocess.PIPE,
+        stderr=subprocess.PIPE,
+        text=True,
+        timeout=600,
+    )
+    if not kubectl_ai_result.returncode == 0:
+        raise Exception(
+            f"kubectl-ai command output. Error: {kubectl_ai_result.stdout.strip()}"
+            f"kubectl-ai command failed. Error: {kubectl_ai_result.stderr.strip()}"
+        )
+    return kubectl_ai_result.stdout.strip()
+
+
 def analyze_failed_tests() -> str:
     """Analyze the failed tests and provide fixes"""
     if Path("rerun-allure-report").exists():
@@ -205,7 +260,7 @@ def analyze_failed_tests() -> str:
             {"role": "system", "content": debug_prompt},
             {"role": "user", "content": "analyse the failed tests"},
         ]
-        payload = {"model": "gemma4:e4b", "messages": messages, "temperature": 0}
+        payload = {"model": "qwen3.5:2b", "messages": messages, "temperature": 0}
         headers = {"Content-Type": "application/json"}
         url = "http://localhost:11434/v1/chat/completions"
         response = requests.post(url, json=payload, headers=headers)
@@ -218,23 +273,39 @@ def analyze_failed_tests() -> str:
 
 def run_test_failure_analysis():
     if os.getenv("PR_ERROR_MSG") == "Failed to Prepare CI environment":
-        response = analyze_env_setup_failure()
+        try:
+            setup_helm_chart(service="kubectl-ai")
+            process = setup_port_forwarding(service="kubectl-ai")
+            assert "qwen3.5:2b" in str(validate_ollama_model())
+            # response = analyze_env_setup_failure()
+            response = analyze_env_setup_failure_using_kubectl_ai()
+        except Exception as e:
+            logger.info(
+                f"Failed to run analyze_env_setup_failure_using_kubectl_ai: {e}"
+            )
+        finally:
+            uninstall_helm_chart(service="kubectl-ai")
     else:
-        response = analyze_failed_tests()
+        try:
+            setup_helm_chart(service="ollama")
+            process = setup_port_forwarding(service="ollama")
+            assert "gemma4:e4b" in str(validate_ollama_model())
+            response = analyze_failed_tests()
+        except Exception as e:
+            logger.info(f"Failed to run analyze_failed_tests: {e}")
+        finally:
+            uninstall_helm_chart(service="ollama")
     if response is None:
         return "No logs found to analyze"
     data = json.loads(response.decode("utf-8"))
     reasoning = data["choices"][0]["message"].get("content")
-    return reasoning
+    return reasoning, process
 
 
 if __name__ == "__main__":
     process = None
     try:
-        setup_ollama_helm_chart()
-        process = setup_port_forwarding()
-        assert "gemma4:e4b" in str(validate_ollama_model())
-        response = run_test_failure_analysis()
+        response, process = run_test_failure_analysis()
         with open("logs/failure_analysis.txt", "w") as f:
             f.write(response)
     except Exception as e:
@@ -243,4 +314,3 @@ def run_test_failure_analysis():
         if process and process.poll() is None:
             process.terminate()
             process.wait()
-    uninstall_ollama_helm_chart()
diff --git a/gen3-integration-tests/gen3_ci/scripts/setup_ci_env.sh b/gen3-integration-tests/gen3_ci/scripts/setup_ci_env.sh
index 21a4bba72..6aa169ce2 100755
--- a/gen3-integration-tests/gen3_ci/scripts/setup_ci_env.sh
+++ b/gen3-integration-tests/gen3_ci/scripts/setup_ci_env.sh
@@ -557,11 +557,14 @@ ci_es_indices_setup() {
 }
 
 wait_for_pods_ready() {
-  export timeout=1800
+  export timeout=600
   export interval=20
 
-  end=$((SECONDS + timeout))
-  while [ $SECONDS -lt $end ]; do
+  SECONDS = 0
+
+  end=$(( $(date +%s) + timeout ))
+  while [ "$(date +%s)" -lt "$end" ]; do
+    echo "Running at $(date)"
     # Get JSON for not-ready, non-terminating pods
     not_ready_json=$(kubectl get pods -l app!=gen3job -n "${namespace}" -o json | \
       jq '[.items[]