Build an AI search engine

Project Title: South African Legal Judgments Semantic Search Engine

Project Objective: To design and implement a standalone semantic AI driven search engine that can retrieve highly accurate and relevant information from a collection of more than 1000 South African legal law judgments, as published by the South African courts. ( See document links below) The search engine should be able to understand the meaning and context of user queries and return results accordingly.

Functional Requirements:
The search engine should be able to handle free text queries entered by users.
The search engine should be built using AI technologies, preferably leveraging natural language processing (NLP) and machine learning algorithms to optimize for accuracy and relevance.
The search engine should be able to search through a collection of over 1000 legal law judgments, as published by the South African courts, in PDF format.
The search engine should be a standalone product, meaning it will not be integrated into an existing application or platform.
The search engine should be capable of handling English-language queries only.
The search engine should be able to retrieve highly accurate and relevant information based on the meaning and context of user queries.
The search engine should be able to handle multiple search queries concurrently and provide quick responses.

Non-Functional Requirements:
The project must be completed within one week.
The budget for the project is $100 USD.
There are no specific data privacy or security requirements for the project.
The search engine will be handed over to a third party for maintenance and future development.

Testing and Evaluation:
The search engine should be evaluated based on mostly accuracy, relevance, and to a lesser degree response time.
The client should be provided with documentation outlining how to use and maintain the search engine and how to add extra documents to the dataset if required.

Development Tools:
The freelancer should use AI technologies, preferably natural language processing (NLP) and machine learning algorithms.
The freelancer is free to choose their preferred programming languages, frameworks, and tools for the project, provided that they meet the project requirements and constraints.

Deliverables:
Source code for the semantic search engine.
Documentation outlining how to use and maintain the search engine.
Test cases and data sets used to evaluate the search engine.
A report summarizing the results of the evaluation and testing of the search engine.
A functional example of the search engine hosted on a mutually agreed host.
The functional example should have a page called "Search" with a single search bar. (See example image attached)
The search bar should accept user input in the form of search terms, phrases, or natural language queries.
When the user submits a search query, the search engine should produce highly relevant results based on the search input and display them in a list.
The functional example should demonstrate the search engine's ability to understand the meaning and context of user queries and return results accordingly.
The functional example should be tested to ensure that it meets the project requirements and constraints.
Any other deliverables agreed upon between the freelancer and the client.

Document links:
http://www.saflii.org/za/cases/ZASCA/2022/
http://www.saflii.org/za/cases/ZASCA/2021/
http://www.saflii.org/za/cases/ZASCA/2020/
http://www.saflii.org/za/cases/ZASCA/...<year>/<number>