Create a simple ASR system

You are asked to implement an ASR system that receives a recording at a time, which is a sentence of 4-10 digits that has been said to have sufficiently long pauses. The system proceeds to segment the proposal and then recognizes each word. The output produces text with the digits identified. • Emphasize the signal processing before starting the recognition phase (eg with appropriate filters, sampling rate change, etc.). • It is important to describe the system algorithmically (attribute extraction, recognition algorithm) and explain its performance, so except for the source code I also expect a super analytical document. This is a very simple task, so please do not overprice it ! 120 EUR is the most acceptable price.