I will build full stack ai platforms with comfyui apis and automation


About this gig
Turn your AI idea or ComfyUI workflow into a web platform your customers can use.
I build frontend UI/UX, backend, APIs, business logic and GPU deployment. My projects include AI photobooth, photo restoration, video upscaling and student-model training.
Services:
- ComfyUI and Python inference workflows
- Responsive web apps and admin dashboards
- APIs, job queues, progress tracking and storage
- GPU server or serverless deployment on Modal or RunPod
- Automation and white-label tools for agencies
- Training or fine-tuning through a custom offer
Basic: one workflow and API. Standard: a focused web MVP. Premium: a larger app with login and admin. Source code and setup documentation are included.
Portfolio examples:
genbooth.vn | restore-img.com | jetthuang.com
Cloud/GPU usage, paid APIs/models and ongoing maintenance are separate. Training depends on the model, dataset, GPU budget and evaluation goals.
Message me on Fiverr with your use case before ordering so we can agree on scope, hosting and delivery. All communication and orders stay on Fiverr.
Get to know Nhat Hoang
Full Stack AI Developer for Workflows, Web Apps and Automation
- FromVietnam
- Member sinceFeb 2025
- Last delivery1 year
Languages
Vietnamese, English
My Portfolio
FAQ
What do you mean by a full stack AI solution?
I connect the user interface, backend, APIs, AI workflow and deployment. Each package has a defined scope. Complex SaaS products, multi-tenant agency systems and model training need a custom offer.
Can you turn my ComfyUI workflow into a web app?
Yes. I can connect an agreed workflow to an API and web interface, with job status and results. Share your workflow, custom nodes, model requirements and sample inputs so I can check compatibility.
Do you offer GPU servers and serverless deployment?
Yes. We choose one deployment target based on model compatibility, workload and budget. Modal and RunPod are options. Serverless does not guarantee zero cold starts or lower costs for every workload.
Is model training or fine-tuning included?
It is available through a custom offer. We define the base model, dataset, training method, GPU budget and evaluation criteria first. The listed packages cover inference integration and do not include a new training run.
Can you work with agencies and white-label products?
Yes. I can develop branded AI tools and integrate with your delivery process. Multi-tenant permissions, client billing and reseller features require an agreed custom scope. I deliver as an independent developer.
Are hosting and AI usage costs included?
No. Cloud hosting, GPUs, storage, paid models and third-party APIs are paid through your own accounts. I identify the required services before implementation; actual usage costs depend on your workload.
Will I receive the source code?
Yes. You receive the custom application code and setup documentation included in your order. Third-party models, libraries and assets remain subject to their own licenses.
What counts as a revision?
A revision is one consolidated round of adjustments within the agreed requirements. New features, additional workflows, changed model choices or deployment targets need a new scope. Defects are assessed against agreed acceptance criteria.
Can you build automation and business logic flows?
Yes. I can connect APIs, job queues, approval steps, progress updates and output handling. We define the inputs, decisions, permissions and error cases first. The package limits still apply; larger integrations need a custom offer.
What do you need before starting?
Your use case, must-have features, existing workflow or code, sample inputs, expected outputs, design references and deployment preferences. For training, include dataset details and usage rights. Send the brief through Fiverr.

