Technical Debt: Improve Topic Search with Accent-Insensitive Matching

## Problem
Currently, our topic search doesn't handle accent-insensitive matching. Searching for "politica" (without accent) doesn't match the existing topic "política" (with accent), leading to poor user experience, especially for Portuguese content where accents are common.

## Path to Explore
We need a solution that:

- Performs accent-insensitive matching at the database level
- Scales efficiently as the topics collection grows
- Doesn't load the entire collection into memory
- Maintains current functionality (limit, sorting, alias matching)

## Potential approaches to investigate:

- MongoDB collation settings for accent-insensitive queries
- Text indexes with language-specific configurations
- Normalized field strategy (storing pre-processed versions)
- MongoDB Atlas Search capabilities (if applicable)

## Impact
- User Experience: Users must type exact accents to find topics, which is cumbersome and error-prone
- Scope: Primarily affects Portuguese language searches where accents are common