[BUG]: Inconsistent model config, hardcoded API endpoints, and fragile logging/search handling

### Bug Description

There are multiple issues in the current pipeline:

-> Different modules use different LLM models, causing inconsistent results
-> Frontend API calls are hardcoded instead of using environment variables
-> Logging has encoding issues (especially on Windows)
-> Fact-check search does not properly handle invalid or empty Google API responses

This makes the system harder to maintain, debug, and deploy across environments.

### Steps to Reproduce

-> Run different analysis modules (bias, sentiment, fact-check, chat)
-> Notice inconsistent outputs due to different model usage
-> Run frontend without proper environment variables
-> Check logs on Windows system
-> Trigger a failed or empty Google search API response
-> Observe errors or crashes in fact-check flow

### Logs and Screenshots

_No response_

### Environment Details

_No response_

### Impact

High - Major feature is broken

### Code of Conduct

- [x] I have joined the [Discord server](https://discord.gg/hjUhu33uAn) and will post updates there
- [x] I have searched existing issues to avoid duplicates