diff --git a/docs/csv-adapter/compiler-layers.md b/docs/csv-adapter/compiler-layers.md
new file mode 100644
index 00000000..c1b3b0ff
--- /dev/null
+++ b/docs/csv-adapter/compiler-layers.md
@@ -0,0 +1,51 @@
+## Schema Onboarding and Ingestion
+0. Config Parser
+  - Define types for a config
+  - Verify if a config is correct on not using Zod
+  - Verify if the file names are in the correct format using the folder mentioned in the config
+  - Verify if the namespaces and folder details are in the right format
+  - Check for spelling errors
+    - Suggest incorrect configurations/spellings
+  - Suggest in case of unused configurations
+  - Error Handling
+    - Raise Issues when this doesn't happen
+    - Defined Errors with Codes
+  - Add a config to silent the errors and report it at the end for incorrect configurations
+1. CSV Format Definition as Types
+  - Dimension
+  - Event
+  - Dimension Data
+  - Event Data
+  - Check the validity of Blacklisted and Whitelisted Dimension combinations
+2. CSV Reader Utilities
+  - Ability to read CSV and check if the format is correct or not
+  - Load the CSV Data using the Types defined in 1
+  - Error Handling
+    - Raise Issues when this doesn't happen
+    - Defined Errors with Codes
+  - Loading and Verifying data based on the schema
+    - Ability to load data files using a dataframe
+    - Ability to verify data formats using a dataframe
+    - Ability to filter null, NA, empty values
+3. Validate Event Grammar to Dimension Grammar Mapping
+  - A comprehensive test suite to test out all the fields in Event getting mapped to a dimension and non dimesnion
+  - Error Handling
+    - Raise Issues when this doesn't happen
+    - Defined Errors with Codes
+4. CSV Format to Dimension Grammar Generator
+  - Ability to generate Ajv Schema as well
+  - Ability to generate types directly
+5. CSV Format to Event Grammar Generator
+  - Ability to generate Ajv Schema as well
+  - Ability to generate types directly
+6. Manage a Global State of steps at which the parer is at
+  - Steps 
+    - 1/N for expected Dimension, Events and Dataset Grammars to be ingested
+    - 1/N for expected Dimension and Dataset Data to be ingested
+    - 1/N for global steps that need to be done for total ingested data
+    - Global list of things that need to be done
+7. Ability to pause and resume the schema ingestion pipeline
+
+## Query Builder
+1. Ability to manage retries
+2. Adapter approach to connect to other databases.
\ No newline at end of file