diff --git a/CHANGES.md b/CHANGES.md
index 71438ae..667d028 100644
--- a/CHANGES.md
+++ b/CHANGES.md
@@ -1,5 +1,14 @@
 # Backend internal interview (rust)
 
-## CHANGES: improved branch 
+## CHANGES: improved branch
 
-  
+- Create abstractions / models to load and parse metrics, can be easily extended 
+- Define basic operations via [`MetricParser`] trait path **defs/metrics.rs**
+- Use **structs** to group metrics attributes and implement operations like [`metrics_consumer::MetricsConsumerTarget`]
+- Use **enums** with associated values like [`defs::metric_data::MetricsConsumerData`] path  **defs/metric_data** to group attributes and values
+- Remove **Maps collections**, use [Vectors](https://doc.rust-lang.org/std/vec/struct.Vec.html)
+- Use **const** for DEFAULT values on [`defs`] module 
+- Remove all <i>hardcoded</i> values to [`Config`] settings, a **TOML config file** (see config.toml) can be used to define several metrics input and customize output
+- **Output** write to file (append o rewrite mode) is working (for now I/O is sync)
+- **Command line** arguments are processed [`parse_args`] with [`defs::CliSettings`] via [clap](https://docs.rs/clap/latest/clap/) and **overide** the ones loaded in **config files**
+- **Tests** have been accommodated from previous [Branch basic-fixed](https://repo.jesusperez.pro/NewRelic/be-technical-interview-rust/src/branch/basic-fixed) version to **imporved** approach
diff --git a/README.md b/README.md
index 53ebd98..b48cd4f 100644
--- a/README.md
+++ b/README.md
@@ -1,21 +1,66 @@
 # Backend internal interview (rust) - improved  
 
+This **Improved** branch is a rather disruptive approach to the [initial proposal](https://repo.jesusperez.pro/NewRelic/be-technical-interview-rust)<br>
+[Branch basic-fixed](https://repo.jesusperez.pro/NewRelic/be-technical-interview-rust/src/branch/basic-fixed) tried to solve proposal from initial code 
+as a **continuity effort** with the necessary changes and some improvement adjustments such as the **parallel input processing**
+
 > [!NOTE]
-> It is able to **process multiple metrics in parallel**. <br>
 > A full refactoring done for <u>better quality, maintenance and be more readable</u>. (Structs, implementaitions, settings for multiple inputs, etc). <br>
+> It is able to **process multiple metrics in parallel**. <br>
+
+## In summary
+
+- [x]  Define a basic model, easily to extend and modify. **Abstraction / Generic**.
+- [x]  Structs and implementations to specific metricis traitments. **Modular appoach**.
+- [x]  Settings and configuration for interactive and non interactive processing (batch mode) **Customize on context**.
 
 See [main changes](/NewRelic/be-technical-interview-rust/src/branch/improved/CHANGES.md)
 
-A **Refactor metric-consumer** task
+> [!IMPORTANT]  
+> Use [ABOUT](ABOUT.md) content for quick [Verification](ABOUT.md) (it requires download and build)
+
+## Benefits
+
+Elements items come from  [main changes](/NewRelic/be-technical-interview-rust/src/branch/improved/CHANGES.md)
+
+| Element          | Benefit |
+|----------------- |---------|
+| Generic traits   | Group generic task for metric processing, steps / tasks separation |
+| Structs          | Customize atributes and implementation for specific target or patterns  |
+| Enums with values| Associate attributes metrics and values, easy to add new attributes or combine with different values at once |
+| Vectors          | Simplify types / grouped in structs, priorize vectors type, easy to iterate, filter, sort, etc |
+| Const and Config | Group main const, define metric targes and operations in declarative mode for non intective traitment |
+| Command line args| Help to run in terminal as a cli |
+| Unit Tests       | Verify some operations results |
+
+> [!TIP]  
+> After download repository and build:
+
+> - use [howto](howto.md) for command, options, etc.
+> - Build documentation and browse content whit source code (instruction in [about](ABOUT.md))
+
+
+## Ideas not included
+
+- **Async I/O** to scale and performance ?
+- Other Thread alternatives like [Tokio](https://tokio.rs/) or/and [Coroutines](https://doc.rust-lang.org/std/ops/trait.Coroutine.html)
+- Benchmarking for optimization
+- More **tests**
+- Run as **API mode** not only as batch processing
+<br>
+
+> [!NOTE]
+> Code is in the private repository with several varias branches.<br>
+> Link to [branch repository improved](https://repo.jesusperez.pro/NewRelic/be-technical-interview-rust/src/branch/improved)
+
+---
 
 [Home Task exercise description](Home-Task_exercise_interview_Rust.pdf)
 
 There are several branches developed as proposal:
 
-- [Basic fixed one]((/NewRelic/be-technical-interview-rust/src/branch/basic-fixed)
+- [Basic fixed one](/NewRelic/be-technical-interview-rust/src/branch/basic-fixed)
   > A basic code review, fixed with minor and essential changes to work as expected. <br>
   > Tests are included for verification. <br>
 
 - [Improve one](/NewRelic/be-technical-interview-rust/src/branch/improved)
-
-- [Multiple input one](multi-input)