Sentra is a visual AI workflow auditor I built to help identify when human oversight is necessary in AI-powered decision-making systems. I was inspired by the growing risks of unchecked automation in areas like hiring, healthcare, and finance, where AI is increasingly making high-impact decisions without accountability. With Sentra, users input a step-by-step workflow, and the system uses GPT-4 to analyze each step, flag potential risks (low, medium, or high), and provide actionable recommendations. I built the frontend using React and the backend using Python with FastAPI, integrating OpenAI’s API through structured prompts. Sentra also includes features like PDF export, animated loaders, tab-based navigation, and built-in example workflows to make the experience intuitive and professional. I’m proud to have created a tool that doesn’t just surface AI risk—it helps people understand and fix it, making responsible automation more accessible.

Built With

Share this project:

Updates