Frontend: N/A (API-focused, potential integration with React/Vue.js for future UI)
Backend: Flask, FastAPI, SQLAlchemy, FFmpeg, AIOHTTP
Database: MySQL, Google Sheets API (for logging)

Projects Overview

We developed a full-stack API solution to assess the credibility of video and audio content using advanced NLP and AI techniques. The system supports transcription, key phrase detection, and source verification using multiple news APIs. Designed for scale, it handles large files securely and efficiently.

This tool reflects our commitment to combating misinformation through AI. It integrates Whisper, NLP libraries, and fast API communication for accurate, real-time credibility scoring.

Challenges & Constraints

Processing and verifying large multimedia files across multiple news APIs required consistent formatting, performance, and security.

  • Aggregating results from news APIs with different response formats and strict rate limits
  • Handling large file uploads and ensuring secure, asynchronous processing of data

Projects Solution

We built a scalable, API-driven system using Flask, FastAPI, and Whisper for speech-to-text, with NLP pipelines and FFmpeg for media handling.

Asynchronous API calls, Redis caching, and normalized output allowed fast and reliable credibility checks with multi-language and secure file support.