Browse categories
Explore
Fiverr Pro
English
$
USD
I will fix your broken comfyui workflow or custom node errors
Tyler L
About this gig
Your workflow worked once. Now it's a wall of red text, a CUDA error, a missing node, or it silently produces garbage. I fix that.
I specialize in the failures tutorials don't cover:
- torch / torchvision / CUDA version mismatches (including custom node C++ extensions that refuse to compile)
- ComfyUI startup crashes and import errors after updates
- VRAM exhaustion on consumer GPUs (8GB and under is my home turf)
- broken model loading: wrong checkpoints, missing VAE/text encoders, gated HuggingFace repos, corrupt safetensors
- stale workflows: renamed nodes, schema changes, LoRA/ControlNet mismatches across SD1.5 / SDXL / Flux
- "my output doesn't match the tutorial" determinism problems
How it works: you send me the error, your workflow JSON, and your system specs. I reproduce the failure, fix it, and send back the working workflow plus a short report on what broke and why, so it doesn't break again next update.
Get to know Tyler L
Tyler L
Freelance AI Pipeline Engineer
- FromUnited States
- Member sinceJun 2020
- Avg. response time1 hour
Languages
English
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.
Other AI Development Services I Offer
FAQ
What if you can't fix it?
If I can't successfully diagnose it, I cancel before delivery and you don't pay. You keep the diagnosis either way.
Can you fix it live on my machine?
Yes - Premium includes a screen share session if the fix needs your exact environment. Message me first for scheduling.
Do you need my models?
Usually no. Model names and hashes are enough; I test with my own copies or matched stand ins.
