Logic Grid Puzzle Solver Development

I'm seeking a developer to create a logic grid puzzle solver that can efficiently handle puzzles with 3 to 5 categories, each containing 3 to 5 values. The program should accept a list of logic clues and automatically compute and return the unique solution, if one exists.

Key Requirements:
- Input should be in JSON or structured data format, detailing categories, values, and clues.
- The output should be a fully solved logic grid presented as structured JSON.
- The program should be written in Python, though other languages are acceptable.
- The algorithm must support generic puzzles using techniques such as backtracking, constraint propagation, or solvers like Z3.

Additional Features:
- Provide step-by-step hints and partial solving stages.
- Support for natural language clues through parsing predefined patterns.

Ideal Skills and Experience:
- Strong proficiency in Python or similar programming languages.
- Experience with algorithms for logic-based games or AI reasoning.
- Familiarity with parsing techniques for handling natural language clues.
- Ability to implement advanced solving techniques like backtracking or constraint propagation.

If you have a passion for puzzles and a knack for problem-solving, I'd love to see your approach to this project!