• 

  App Interface

Inspiration I was inspired by the challenge of making legal documents accessible to everyday users. Contracts and case files are often filled with jargon, and I wanted to build a tool that could decode complexity into clarity. Gemini’s advanced language capabilities gave me the confidence to tackle this.

What I Learned How to fine‑tune prompts for extracting obligations, risks, and deadlines.

The importance of structured outputs (Markdown tables, clause highlights) for readability.

How AI can balance detail and simplicity, avoiding overwhelming users with too much text.

The need for trust and compliance when analyzing sensitive legal content. How I Built It (with Gemini) Frontend: A React interface with a paste/upload box for contracts or case files.

Backend: Firebase for authentication and secure storage.

AI Layer (Gemini):

Used Gemini to parse text into clause categories (termination, liability, confidentiality).

Applied Gemini’s summarization to generate plain‑language explanations.

Designed risk scoring with a simple math model:

𝑅

∑

𝑖

1 𝑛 𝑤 𝑖 ⋅ 𝑐 𝑖

= clause risk rating. Output: Markdown‑styled analysis with highlights, summaries, and risk scores. Challenges I Faced Ambiguity in legal text: Gemini sometimes struggled with vague clauses. Data privacy: Ensuring uploaded contracts were handled securely. Hackathon time limits: Building a working prototype quickly while keeping quality. UX optimization: Making results clear and actionable without overwhelming users.

What It Does Summarizes contracts into obligations, rights, and risks. Flags red‑flag clauses (e.g., unilateral termination, hidden fees). Suggests improvements for clarity and fairness. Scores risk using clause‑based models. Delivers structured analysis in Markdown for easy sharing.

Built With

• aistudio