Part-Time AI & Full-Stack Help
Budget / SalaryHourly project
TypeFreelance project
LocationRemote
Posted1 hour ago
I’m looking for someone who can split their time between two intertwined needs.
First, I need a supervised-learning model built from the ground up that can handle time-series prediction. I already have historical data; what I’m missing is the full model-development workflow—feature engineering, training, validation, and iteration—implemented in clean, well-documented Python (TensorFlow, PyTorch, or similar). Accuracy, reproducibility, and the ability to improve the model later are essential.
Second, my web platform (Node.js + React on a Postgres backend) needs ongoing full-stack enhancements so the forecasting engine can be surfaced through a secure API and intuitive dashboard. You’ll take model outputs, expose them through endpoints, and build the front-end views that let users explore forecasts in real time.
Key deliverables:
• A version-controlled repo containing the complete time-series prediction pipeline
• REST (or GraphQL) endpoints that serve model results and accept new data for retraining
• Front-end components that visualize forecasts and basic performance metrics
I expect a part-time cadence—roughly 10–15 hours a week—with milestones we’ll define together. If you’re comfortable jumping between deep learning notebooks and React components, let’s talk.
First, I need a supervised-learning model built from the ground up that can handle time-series prediction. I already have historical data; what I’m missing is the full model-development workflow—feature engineering, training, validation, and iteration—implemented in clean, well-documented Python (TensorFlow, PyTorch, or similar). Accuracy, reproducibility, and the ability to improve the model later are essential.
Second, my web platform (Node.js + React on a Postgres backend) needs ongoing full-stack enhancements so the forecasting engine can be surfaced through a secure API and intuitive dashboard. You’ll take model outputs, expose them through endpoints, and build the front-end views that let users explore forecasts in real time.
Key deliverables:
• A version-controlled repo containing the complete time-series prediction pipeline
• REST (or GraphQL) endpoints that serve model results and accept new data for retraining
• Front-end components that visualize forecasts and basic performance metrics
I expect a part-time cadence—roughly 10–15 hours a week—with milestones we’ll define together. If you’re comfortable jumping between deep learning notebooks and React components, let’s talk.
Apply on Freelancer →
Project sourced from Freelancer.com. Applications happen directly on the original platform — we never collect your data.