Cross-Platform Prediction Bot Development

I want to build a prediction bot that works seamlessly on the web, mobile, and desktop. The heart of the project is automated data analysis: the system should pull fresh data from the sources we define, process it with a machine-learning model, and publish clear, up-to-date predictions without any manual touch-points. To make sure the bot is portable, I’m open to the tech stack you feel is best—Python with TensorFlow or PyTorch for the model, a Node.js or FastAPI back end, and a front end that can compile to the browser, iOS/Android, and Windows/macOS (React, React Native, Flutter, or Electron are all fine). Stability, maintainability, and clean documentation matter more to me than the exact framework. Deliverables • Architecture and stack proposal geared toward web, mobile, and desktop deployment • Data pipeline that ingests, cleans, and stores incoming data automatically • Trained or integrated ML model that produces reliable predictions on a schedule (or in real time if feasible) • Lightweight UI for each platform showing current and historical prediction outputs • Clear build/run instructions, plus guidance on retraining or swapping the model later Acceptance criteria 1. One codebase (or tightly coupled modules) that deploys to all three platforms. 2. End-to-end automation from data retrieval through to on-screen prediction. 3. Unit tests covering key data-handling and prediction logic. Please share examples of past predictive-analytics projects, the libraries you prefer for cross-platform work, and a rough timeline. I’m ready to get started as soon as we align on the approach.