diff --git a/README.md b/README.md
index 6ded9648..db94fc60 100644
--- a/README.md
+++ b/README.md
@@ -1 +1,46 @@
-# atoma-node
+# Atoma Node infrastructure
+
+## Introduction
+
+The present repository contains the logic to run an Atoma node. Atoma nodes empower the Atoma Network, a decentralized network
+for verifiable AI inference. Nodes can lend their GPU compute to the protocol, so large language models (LLMs for short) can be
+run across a decentralized network, at a cheaper cost. Moreover, nodes can be organized in order to provide verifiability guarantees
+of the correctness of their generated outputs. This means that the Atoma Network can empower smart contracts, deployed on any blockchain,
+to run verifiable inference and guarantee an intelligence layer to Web3. 
+
+## Run a node
+
+In order to run an Atoma node, you should provide enough evidence of holding a powerful enough machine for AI inference. We
+allow a wide range of possible hardware, including NVIDIA GPU series (RTX3090, RTX4090, A100, etc), as well as a good enough
+Internet bandwidth connection. We further highly encourage nodes to register on Atoma contracts on supported blockchains
+(these include Arbitrum, Solana, Sui, etc). In order to register, we suggest the reader to follow the instructions in the
+Atoma contract [repo](https://github.com/atoma-network/atoma-contracts).
+
+Once registration has been completed, the user is required to:
+
+1. Clone this repo (it is assumed that Rust is installed, otherwise follow the instructions [here](https://www.rust-lang.org/tools/install)).
+2. Create configuration files for both the model inference service, the event listener service and the blockchain client service. Notice
+that the event listener and blockchain client services need to be for the same blockchain, in which the node has previously registered. That said, a single node can be registered in multiple blockchains and listen to events on each of these (for higher accrued rewards).
+3. The model inference service follows schematically (in toml format):
+
+```toml
+api_key = "<YOUR_HUGGING_FACE_API_KEY>" # for downloading models
+cache_dir = "<PATH_FOR_MODEL_STORAGE>" # where you want the downloaded models to be stored
+flush_storage = true # when the user stops the Atoma node, it flushes or not the downloaded models
+jrpc_port = 3000 # Atoma node JRPC port
+models = [[device, precision, model_type, revision, use_flash_attention], ...] # Specifications for each model the user wants to operate, as an Atoma Node
+tracing = true # Allows for tracing
+```
+4. The event subscriber service configuration file is specified as (in toml format):
+
+```toml
+http_url = "RPC_NODE_HTTP_URL" # to connect via http to a rpc node on the blockchain
+ws_url = "RPC_NODE_WEB_SOCKET_URL" # to connect via web socket to a rpc node on the blockchain, relevant for listening to events
+package_id = "SUI_PACKAGE_ID" # the Atoma contract object id, on Sui.
+small_id = 28972375 # a unique identifier provided to the node, upon on-chain registration
+
+[request_timeout] # a request timeout parameter
+secs = 300
+nanos = 0
+
+```
\ No newline at end of file
diff --git a/atoma-types/src/lib.rs b/atoma-types/src/lib.rs
index 895e340f..ad0f6eb3 100644
--- a/atoma-types/src/lib.rs
+++ b/atoma-types/src/lib.rs
@@ -4,6 +4,14 @@ use serde_json::{json, Value};
 
 pub type SmallId = u64;
 
+/// Represents a request object containing information about a request
+/// Prompt event, emitted on a given blockchain, see https://github.com/atoma-network/atoma-contracts/blob/main/sui/packages/atoma/sources/gate.move#L45.
+/// It includes information about a ticket ID, sampled nodes, and request parameters.
+///
+/// Fields:
+/// id: Vec<u8> - The ticket ID associated with the request (or event).
+/// sampled_nodes: Vec<SmallId> - A vector of sampled nodes, each represented by a SmallId structure.
+/// body: serde_json::Value - JSON value containing request parameters.
 #[derive(Clone, Debug, Deserialize)]
 pub struct Request {
     #[serde(rename(deserialize = "ticket_id"))]
@@ -61,6 +69,9 @@ impl TryFrom<Value> for Request {
     }
 }
 
+/// Parses the body of a JSON value. This JSON value is supposed to be obtained
+/// from a Sui `Text2TextPromptEvent`,
+/// see https://github.com/atoma-network/atoma-contracts/blob/main/sui/packages/atoma/sources/gate.move#L28
 fn parse_body(json: Value) -> Result<Value> {
     let output = json!({
             "max_tokens": parse_u64(&json["max_tokens"])?,
@@ -90,6 +101,8 @@ fn parse_f32_from_le_bytes(value: &Value) -> Result<f32> {
     Ok(f32::from_le_bytes(f32_le_bytes))
 }
 
+/// Parses an appropriate JSON value, representing a `u64` number, from a Sui
+/// `Text2TextPromptEvent` `u64` fields.
 fn parse_u64(value: &Value) -> Result<u64> {
     value
         .as_str()
@@ -98,6 +111,12 @@ fn parse_u64(value: &Value) -> Result<u64> {
         .map_err(|e| anyhow!("Failed to parse `u64` from string, with error: {e}"))
 }
 
+/// Represents a response object containing information about a response, including an ID, sampled nodes, and the response data.
+///
+/// Fields:
+/// id: Vec<u8> - The ticket id associated with the request, that lead to the generation of this response.
+/// sampled_nodes: Vec<SmallId> - A vector of sampled nodes, each represented by a SmallId structure.
+/// response: serde_json::Value - JSON value containing the response data.
 #[derive(Debug)]
 pub struct Response {
     id: Vec<u8>,