Review intro.md

docs/articles/intro.md

@@ -4,10 +4,10 @@ Efficiently managing large data spans is crucial for achieving optimal performan
 
 However, implementing these optimizations may introduce complexity to the codebase. Striking a balance between performance and maintainability is challenging, especially when optimizations are scattered across various methods. This library addresses this challenge by providing reusable and high-performance abstractions specifically designed for operations on data spans.
 
-This library is inspired on `System.Numerics.Tensors` but extending it to all value types.
+This library takes inspiration from `System.Numerics.Tensors` but widens its scope to include all value types. It offers extra operations and provides an API to support the development of custom operations.
 
-The library seamlessly integrates parallelization and vectorization for optimal performance whenever applicable. Explore the subsequent sections to understand its usage and learn how to tailor it to your specific needs.
+The library seamlessly integrates CPU parallelization and vectorization for optimal performance whenever applicable. Explore the subsequent sections to understand its usage and learn how to tailor it to your specific needs.
 
 The degree of performance enhancement depends on factors such as data type, vectorization potential, availability, and vectorization size. Check out the benchmarks section to assess potential performance gains tailored to your unique use case.
 
-Enhance the performance of your projects without requiring expertise in low-level CPU features.
+Enhance the performance of your projects without requiring expertise in low-level CPU features.