Parcourir les catégories
Explorer
Fiverr Pro
Français
$
USD
I will build a custom comfyui workflow that runs reliably on your GPU
Tyler L
À propos de ce service
Anyone can wire nodes together. I build pipelines that still work on the 50th run tested, documented, and sized to the VRAM you actually have.
I've built and run pipelines for: batch character generation with identity locking (face stays consistent across hundreds of images), inpainting-based wardrobe variation, image-to-3D asset generation with PBR texturing, and fully headless automation through ComfyUI's HTTP API.
What "production-ready" means here:
- Runs on YOUR hardware. I develop on an RTX 3060 Ti 8GB if it runs for me, it runs for you. No "works on my A100" surprises.
- Comes with a runbook: exact versions, model files, install steps.
- Batch-ready: process 10 images or 10,000 with the same graph, with failure logging so one bad input doesn't kill the run.
- API-driveable if you need it: trigger generations from your own scripts or app, not just the web UI.
Every delivery is verified by actually running your pipeline end-to-end before handoff. You get the workflow JSON, the runbook, and sample outputs.
Every order of basic, standard, or premium, comes with 2 complimentary revisions. See FAQ
Découvrez Tyler L
Tyler L
Freelance AI Pipeline Engineer
- DeÉtats-Unis
- Membre depuisjuin 2020
- Temps de réponse moy.1 heure
Langues
Anglais
I am a freelance AI pipeline engineer specializing in local generative-AI systems, including ComfyUI workflow development and image-to-3D asset production. I build and repair pipelines that run reliably outside of cloud infrastructure, focusing on version-pinned and VRAM-aware solutions. With a background in technical troubleshooting and corporate training, I deliver systems complete with documentation and client instruction.
When you choose me, you choose a guy who values his clients, and you'll get a personal one on one connection and feel like your problems matter.
Autres services de Développement IA I Offre
FAQ
Can you match a Youtube/Tutorial workflow that no longer works?
Yes! Send the link, and i'll rebuild it against current ComfyUI and current model versions, and note what changed.
Do I need to buy anything else?
No. Everything I build runs locally on free/open models. Your only cost is your own hardware and electricity.
Can you train a LoRA as part of the pipeline?
Yes! That lives in the Add-On section. $50-100 depending on the dataset size. Message me first for a quote.
How many revisions are included?
Two consolidated revision rounds per order. Send changes as one list; further rounds are quoted as a small add-on
