NewsCaster.AI

Inspiration

We wanted to make news more accessible, personalized, and engaging. Traditional platforms overload users with information, so we set out to create an AI newscaster that delivers only what matters: summarized, clear, and tailored to each person.

What it does

NewsCaster.AI generates real-time, personalized news briefings in a custom voice and style. Users pick their preferred tone, avatar, and topics, and the system delivers dynamic audio or video summaries that feel like a personal news anchor.

How we built it

We combined Next.js for the frontend, Flask for backend logic, and multiple APIs including NewsAPI for real-time data. We integrated OpenAI for summarization and script generation, D-ID for video avatar generation, and Fish Audio for realistic TTS.

Challenges we ran into

Integrating multiple APIs smoothly was difficult, especially syncing audio, video, and generated scripts. Managing rate limits, formatting outputs, and ensuring consistent avatar lip-syncing also took time. Designing a seamless user experience across tools was another challenge.

Accomplishments that we're proud of

We successfully built a full pipeline that turns raw news into a polished, personalized newscast. Our custom avatars, smooth summaries, and clean UI came together exactly how we envisioned. Seeing the first working video output felt huge.

What we learned

We learned how to integrate AI media tools, manage asynchronous workflows, and structure a multi-service system. We also gained experience handling API limitations, improving prompt engineering, and refining UX for non-technical users.

What's next for NewsCaster.AI

We plan to add multilingual anchors, user accounts with saved preferences, more expressive avatars, and support for custom feeds like sports, finance, or local news. Eventually, we aim to launch a mobile app and expand into fully interactive AI broadcasts.