diff --git a/docs/deployment_design/components.plantuml b/docs/deployment_design/components.plantuml
new file mode 100644
index 0000000..92df7d0
--- /dev/null
+++ b/docs/deployment_design/components.plantuml
@@ -0,0 +1,24 @@
+@startuml
+left to right direction
+interface REST
+
+component "LlamaIndex API" as api {
+    package app as "flask app"
+    package LlamaIndex as "llamaIndex_db" {
+        package llama_index
+    }
+}
+app --> LlamaIndex: uses
+REST -- api
+
+folder Container as "Index folder" {
+    file doc as "default__vector_store.json"
+    file index as "docstore.json"
+    file vector as "graph_store.json"
+    file graph as "image__vector_store.json"
+    file image as "index_store.json"
+}
+
+llama_index --> Container : reads
+
+@enduml
diff --git a/docs/deployment_design/components.png b/docs/deployment_design/components.png
new file mode 100644
index 0000000..0939b02
Binary files /dev/null and b/docs/deployment_design/components.png differ
diff --git a/docs/deployment_design/deployment.plantuml b/docs/deployment_design/deployment.plantuml
new file mode 100644
index 0000000..fd25aaf
--- /dev/null
+++ b/docs/deployment_design/deployment.plantuml
@@ -0,0 +1,16 @@
+@startuml
+left to right direction
+interface REST
+
+cloud Azure {
+    node docker as "Docker container" {
+        component "LlamaIndex API" as api
+    }
+    database volume as "Docker Volume" {
+        folder Container as "Index folder"
+    }
+}
+
+REST -- api
+api --> Container : reads
+@enduml
diff --git a/docs/deployment_design/deployment.png b/docs/deployment_design/deployment.png
new file mode 100644
index 0000000..e62dd3f
Binary files /dev/null and b/docs/deployment_design/deployment.png differ
diff --git a/docs/deployment_design/design.md b/docs/deployment_design/design.md
new file mode 100644
index 0000000..be94ff2
--- /dev/null
+++ b/docs/deployment_design/design.md
@@ -0,0 +1,61 @@
+# Llama index deployment design
+
+## Component diagram
+
+![Alt text](components.png)
+
+## Deployment diagram
+
+![Alt text](deployment.png)
+
+## LlamaIndex API startup sequence diagram
+
+```mermaid
+sequenceDiagram
+    participant creator as App Creator
+
+    create participant sc as :StorageContext
+    creator->>sc: create with index_folder
+    create participant index as :BaseIndex
+    creator->>index: create with storage_context
+    create participant retriever as :VectorIndexRetriever
+    creator->>retriever: create with index
+    create participant rqe as :RetrieverQueryEngine
+    creator->>rqe: create with retriever
+    create participant conf as :Config
+    creator->>conf: create with query_engine
+    create participant app as :Flask
+    creator->>app: create with config
+
+```
+
+**Resources:**
+
+- [Loading data in LlamaIndex using a remote
+  filesystem](https://docs.llamaindex.ai/en/stable/module_guides/storing/save_load.html#using-a-remote-backend)
+
+## LLamaIndex API search sequence diagram
+
+```mermaid
+sequenceDiagram
+    actor user
+    participant app as app:Flask
+    participant config as config:Config
+    participant rqe as :RetrieverQueryEngine
+
+    user->>+app: POST /search query_string
+    app->>config: get query_engine
+    app->>+rqe: query with query_string
+    rqe-->>-app: response
+    app->app: generate results from response.nodes
+    app-->>-user: results
+```
+
+**Resources:**
+
+- [Creating a flask API with
+  LlamaIndex](https://docs.llamaindex.ai/en/stable/understanding/putting_it_all_together/apps/fullstack_app_guide.html#flask-backend)
+
+**Note:**
+
+The retriever query potentially takes some time (more than 5 seconds in some instances). This could be problematic as the user expects near instant results. We might have to optimize this process.