Contact person: Generative Analytics Algorithm Development Project

Phone:Show

Email:Show

Location: dubai, United Arab Emirates

Budget: Recommended by industry experts

Time to start: As soon as possible

Project description:
"A diverse collection of structured, unstructured, and semi-structured data is ready for exploration, and I want to push beyond classic descriptive or predictive work into truly generative analysis. The core objective is to design and validate brand-new algorithms that can create insights, content, or synthetic data from these mixed-format sources.

The work will involve:
• Building an end-to-end pipeline that ingests all three data types, handles cleaning, feature engineering, and efficient storage
• Researching and prototyping novel generative techniques—feel free to draw on GANs, diffusion models, transformers, or any other state-of-the-art approach that fits the data
• Iterating until the algorithms meet agreed-upon metrics for quality, diversity, and performance
• Packaging the final solution with well-documented Python code (TensorFlow, PyTorch, or comparable), reproducible notebooks, and a concise technical report that explains model architecture, training logic, and evaluation results

Acceptance criteria will be a repo that runs end-to-end on my environment, produces the stated outputs without errors, and a report that lets a competent data scientist follow your reasoning step by step.

If you thrive on pioneering algorithms and enjoy working across heterogeneous datasets, this project should be a great fit." (client-provided description)