Inspiration
Legal analysis is time-consuming, expensive, and often inaccessible to individuals, small law firms, and organizations with limited resources. I was inspired by the need to make structured legal reasoning more accessible and efficient. The FIRAC methodology is widely used in legal education and practice, yet applying it consistently requires significant effort. LexFIRAC was created to bridge this gap using AI.
What it does
LexFIRAC is a legal case analyzer that applies the FIRAC framework (Facts, Issues, Rules, Analysis, Conclusion) to legal cases. Users can submit a case designation along with evidence or transcript text, and the system generates a structured legal analysis. At each stage of the process, the application can automatically notify the lawyer and/or involved parties with clear, actionable insights.
How we built it
LexFIRAC was built using Google AI Studio, Google Gemini 3, and TypeScript for the user interface, with Gemini 3 serving as the core reasoning engine. I leveraged Gemini 3’s advanced reasoning and long-context capabilities to analyze complex legal texts and generate structured FIRAC-based outputs. I also integrated function calling to automatically send legal analysis results via email to lawyers or relevant stakeholders at key stages of the process. The system follows a modular architecture, enabling scalability and seamless future integrations.
Challenges we ran into
One of the main challenges was ensuring consistency in legal reasoning while minimizing hallucinations. Structuring outputs strictly around the FIRAC framework and defining clear function schemas helped improve reliability. Another challenge was balancing analytical depth with clarity for non-technical users.
Accomplishments that we're proud of
- Successfully implemented structured FIRAC-based legal analysis using Gemini 3
- Integrated function calling for automated legal notifications
- Designed a repeatable and scalable legal analysis workflow
What we learned
I learned how powerful large language models can be when guided by structured legal frameworks. Treating Gemini 3 as an active reasoning agent, rather than a simple text generator, significantly improved trust and usability.
What's next for LexFIRAC - Legal Case Analyzer
Next, I plan to add multi-jurisdiction support, document uploads, confidence scoring, and integrations with legal case management systems. The long-term goal is to make LexFIRAC a reliable AI assistant for real-world legal workflows.
Built With
- ai
- generative-ai
- llm
- typescript
Log in or sign up for Devpost to join the conversation.