AI Keyword Review Popup
Budget: $30 – $250 USD
I’m building a community-management DApp and need a self-contained module that pops up in React, scans the user’s drafted text, and does two things in under one second: (1) suggests helpful keywords and (2) flags anything sensitive. The current sensitivity list includes financial terms such as “investment”, personal information such as “email”, and anything related to political events or politicians; feel free to design the rules so we can expand that list later.
The popup should appear as the user types and show the recommended keywords as a simple list. Your logic layer can be rule-based or a lightweight ML model, but please implement it with TensorFlow.js so it stays entirely in the browser. All tag data needs to flow to MongoDB so the app can query and return fresh suggestions in real time.
Key acceptance criteria
• < 1 s total response time from keystroke to suggestion/flag
• ≥ 90 % keyword recommendation accuracy over 150 simulated test cases that you provide
• Feature branches, daily pull requests, and clear commit messages in our private GitHub repo (I’ll grant access after kickoff)
• Clean README and brief test report summarising methodology and results
The timeline is one week, so strong familiarity with React, TensorFlow.js, and MongoDB is essential. If the above sounds straightforward, let’s get started—I’ll be reviewing PRs daily and can answer questions as they arise.
Budget: $150-250USD
The popup should appear as the user types and show the recommended keywords as a simple list. Your logic layer can be rule-based or a lightweight ML model, but please implement it with TensorFlow.js so it stays entirely in the browser. All tag data needs to flow to MongoDB so the app can query and return fresh suggestions in real time.
Key acceptance criteria
• < 1 s total response time from keystroke to suggestion/flag
• ≥ 90 % keyword recommendation accuracy over 150 simulated test cases that you provide
• Feature branches, daily pull requests, and clear commit messages in our private GitHub repo (I’ll grant access after kickoff)
• Clean README and brief test report summarising methodology and results
The timeline is one week, so strong familiarity with React, TensorFlow.js, and MongoDB is essential. If the above sounds straightforward, let’s get started—I’ll be reviewing PRs daily and can answer questions as they arise.
Budget: $150-250USD