CineShield AI

AI-powered multi-agent system that analyzes screenplays for copyright and trademark risks, explains legal concerns, and generates safer creative rewrites in seconds.

Comment

Inspiration

Film production mein screenplay ki legal clearance process time-consuming aur expensive ho sakti hai. CineShield AI ka idea isi problem se inspired hai: filmmakers aur studios ko screenplay ke andar potential copyright aur trademark risks ko early stage par identify karne mein help karna.

Hum ek aisa AI system banana chahte thay jo sirf risky content ko flag na kare, balki us risk ko explain kare aur safer creative alternatives bhi suggest kare.

What it does

CineShield AI is an AI-powered Studio Legal & Script IP Risk Checker.

Users screenplay files upload kar sakte hain, including PDF, TXT, and Fountain formats. The system parses the screenplay and identifies characters, brands, products, trademarks, dialogue, scene information, and other relevant entities.

A multi-agent pipeline then analyzes the screenplay:

  • Entity Extractor Agent identifies characters, brands, products, trademarks, and character-defining traits.
  • IBM watsonx Auditor Agent performs IP and trademark matching.
  • Gemini Copyright Risk Analyzer evaluates potential conflicts and calculates a 0–100% legal risk score.
  • Findings are categorized as High, Medium, or Low Risk.
  • Gemini Creative Rewriter generates alternative text for flagged content while maintaining the scene's emotional intent and pacing.

The results appear in an interactive screenplay dashboard with color-coded risk highlights, explanations, and rewrite suggestions.

How we built it

CineShield AI was designed as a multi-agent AI workflow connecting screenplay parsing, IP auditing, legal-risk evaluation, and creative rewriting.

The frontend uses a modern web interface with React/Next.js and Tailwind CSS. The backend is designed around Node.js or Python for screenplay parsing and agent orchestration.

The workflow follows a structured sequence: screenplay upload → parsing → entity extraction → IBM watsonx auditing → Gemini risk analysis → interactive risk visualization → creative rewriting.

Challenges we ran into

One of the main challenges was designing a pipeline capable of processing long screenplay files while keeping the different analysis stages organized and connected.

Another challenge was turning complex IP and legal-risk findings into a simple visual experience where filmmakers can quickly understand what was flagged, why it was flagged, and what alternative wording could be considered.

Accomplishments that we're proud of

We are proud of creating a complete concept for an AI-powered screenplay risk-analysis workflow that combines specialized agents instead of relying on a single AI prompt.

We are especially proud of connecting entity extraction, IP matching, risk scoring, visual screenplay highlighting, and AI-powered creative rewriting into one continuous workflow.

What we learned

We learned how multi-agent systems can divide a complex problem into specialized tasks and pass structured information between agents.

We also learned the importance of combining structured screenplay parsing with semantic AI analysis so that risk findings can remain connected to the original screenplay context.

What's next for CineShield AI

Next, CineShield AI can be expanded with stronger IP and trademark data integrations, more advanced long-form screenplay processing, improved agent orchestration, and deeper legal-risk evaluation.

The long-term goal is to make screenplay clearance faster and more accessible while keeping professional legal review as an important part of the filmmaking process.

Built With

Share this project:

Updates