Stable Diffusion / SDXL Expert for LoRA Training — Realistic Male Portraits

We’re looking for a Stable Diffusion / SDXL specialist to improve the consistency and quality of the visual quality and consistency photorealistic images of adult men, including portraits and full-body compositions. The first milestone is to review our existing image-generation setup, model, and LoRA, identify likely causes of inconsistent results, and recommend an approach before starting new training. Further work may include dataset curation and captioning, LoRA training, and integration into our existing workflow. Deliverables should include the trained LoRA, reproducible training and generation settings, a documented workflow, and before-and-after comparisons using agreed prompts and seeds. Evaluation should demonstrate consistent performance across a representative test set, rather than only selected successful images. Please share relevant examples, explain your personal contribution to those projects, and propose a scope and fixed price for the initial review.