Redactify - AI powered Document and Image Redactor

Manually redacting PDFs is tedious, error-prone and doesn't scale. Lawyers, HR teams, healthcare administrators and researchers spend hours hunting through documents to black out sensitive data before sharing them. Redactify provides a solution to this problem by automating this process. Redactify is built on the Fireworks AI API, leveraging open-source large language models accelerated by AMD GPU compute resources. Every document analysis request is processed through Fireworks' AMD-powered inference infrastructure, enabling fast, scalable and reliable extraction of sensitive data without requiring users to manage their own AI hardware. Furthermore, the application combines modern document processing and AI tooling into a single workflow. OpenAI is used to call Fireworks AI's inference endpoint, sending rendered documents and extracted text tokens to a vision model. This results in an application that transforms document privacy from a manual task into an AI assisted process. Powered by Fireworks AI, personally identifiable information (PII) and sensitive data, such as names, emails and phone numbers, are detected directly from the uploaded PDFs and images. Users can then review AI-generated suggestions, adjust redactions as needed and apply them with a single click. Redactify utilises automation while still maintaining human oversight, allowing users to verify detections before finalising the redaction. This combination makes the process faster and more accurate, helping organisations protect confidential information.