Phd technical side

I am a PhD candidate working on a deep learning research project in reciprocal recommender systems. I am attaching a detailed technical plan, but I want to summarize the work here so you can decide whether this project matches your expertise. The project is about building and validating an explainable deep reciprocal recommender system for two-party matching. The model should recommend suitable matches from both sides, not only from one user to one item. The empirical validation will use public reciprocal matching datasets in social matching and recruitment. Affiliate marketing is the motivating case, but the attached plan explains why real public reciprocal datasets are used for validation. I already have a previous code prototype developed by another programmer. Some parts may be reused only after review and validation, such as the two-stage retrieval and ranking idea, embeddings, SentenceTransformer text encoding, FAISS retrieval, and some ranking metrics. However, the data pipeline, label mapping, leakage control, dataset splits, candidate pools, negative sampling, baseline evaluation, reciprocal metrics, result logging, and reproducibility scripts must be rebuilt properly. I am not looking for someone to write my thesis for me. I will handle the academic writing and final research decisions. I need a technical research partner who can discuss the idea with me, review whether the methodology is sound, suggest improvements if needed, and take full responsibility for the technical implementation. The technical work includes data preparation, leakage-controlled splitting, running baselines, building the proposed deep model, implementing feature-level explanations, evaluating with ranking and reciprocal metrics, running ablation studies, saving raw results, and delivering reproducible code and documentation. A key issue in this project is reliability. I do not want unsupported results or hidden problems. If there is any issue with the data, model, baseline, metrics, leakage, or feasibility, I expect you to tell me directly so we can solve it together. Reproducibility is mandatory. The work will not be considered complete unless I can rerun the delivered files and reproduce the reported results. The timeline is tight. I aim to complete the PhD work within approximately two months, so I need someone serious, organized, and experienced with ML research implementation. Please review the attached plan and let me know: 1. Whether you can technically implement this project. 2. Whether you see any weakness or possible improvement in the idea. 3. Your experience with recommender systems, ranking models, FAISS, SHAP. 4. Whether you prefer hourly payment, fixed price, or milestone-based payment. 5. Your estimated timeline and availability. If the collaboration works well, I am also interested in long-term research and technical collaboration, as I work at a university and may need further academic and technical projects in the future.