Senior AI SaaS Engineer / Claude Code Expert — Generative Image & Video Platform

We are building a SaaS platform focused on AI-powered content creation, combining multiple image and video generation technologies into one streamlined product. The platform is already under active development. We are not looking for someone to build a quick prototype from scratch. We need an experienced engineer who can enter an existing codebase, understand the architecture quickly, finish incomplete areas, solve difficult bugs, improve reliability and help us continue developing the product over the long term. We are looking for someone who sits at the intersection of: Senior software engineering + Generative AI + creative technology. This is particularly important to us. We don't want someone who only knows how to connect an API. We want someone who understands why one AI generation is better than another, how image/video generation workflows behave, how prompts and model parameters affect results, how to troubleshoot inconsistent outputs, and how to turn rapidly changing AI technology into a reliable SaaS product. Technologies & areas you'll work with Our platform works with — or will continue integrating — technologies such as: Claude Code / AI-assisted software development Google Gemini / Nano Banana image generation models Seedream Seedance ComfyUI LoRA training and inference Image-to-image generation Text-to-image generation Image-to-video generation Text-to-video generation Video-to-video workflows Reference-image and character-consistency workflows Multiple third-party AI generation APIs Open-source and hosted generative AI models New image/video models as they become available Knowledge of other platforms and models such as Veo, Kling, FLUX and similar technologies is a strong advantage. The AI landscape changes extremely quickly, so actively following new models, APIs, techniques and releases is part of the role. What you'll be doing You will work directly on the existing SaaS application and help take it from its current development stage into a reliable, polished production platform. Your work will include: Understanding and improving an existing codebase Using Claude Code extensively as part of the development workflow Finishing existing features and building new ones Debugging difficult problems across frontend, backend, APIs and infrastructure Integrating new AI image and video generation providers Building clean abstractions around multiple AI providers and models Managing asynchronous generation jobs Handling API errors, failed generations, retries, rate limits, timeouts and provider outages Improving generation status/progress handling Improving reliability, logging and observability Optimizing generation speed and cost Working with ComfyUI workflows and APIs Troubleshooting ComfyUI nodes and custom workflows Creating or improving LoRA training pipelines Working with datasets, captioning, training, evaluation and model versioning Improving prompting and generation logic Testing new models and deciding whether they are worth integrating Helping maintain consistent output quality across different models Identifying technical debt and architectural problems before they become bigger issues Writing maintainable production code rather than relying blindly on AI-generated code Claude Code experience is important We use AI-assisted development heavily. You should be genuinely experienced with Claude Code, including working with larger existing repositories, planning complex changes, debugging, testing and using AI coding agents efficiently. However, we are not looking for "vibe coding" without engineering knowledge. You need to understand the code Claude produces, recognize when an approach is wrong, debug it yourself and make sound architectural decisions. Generative AI experience is equally important You should already be hands-on with modern generative image and/or video workflows. Ideally you have personally worked with several of the following: Nano Banana / Gemini image models, Seedream, Seedance, ComfyUI, LoRA training, FLUX, Veo, Kling or comparable image/video generation systems. You should understand concepts such as: Prompting Reference images Image conditioning Character/style consistency Seeds and reproducibility Image editing Image-to-video Model parameters API limitations Generation queues GPU workflows LoRA training Dataset preparation Model evaluation Workflow optimization Visual judgement matters This is a content-creation product, so having a good eye for visual quality is extremely important. You don't need to be a professional artist, but you should be able to look at two generations and understand why one feels more realistic, consistent or professionally produced than another. Attention to small details is important. If something looks wrong, we expect you to notice it. Who we're looking for You are probably a good fit if you: Have strong professional software-development experience Are comfortable taking ownership of an existing codebase Have built or worked on production SaaS applications Have real experience integrating third-party APIs Understand asynchronous/background processing Have strong debugging and troubleshooting skills Are comfortable working across frontend and backend when necessary Have hands-on Generative AI experience Have worked with image and/or video generation models Understand ComfyUI beyond simply downloading someone else's workflow Have experience with LoRA training or custom model workflows Use Claude Code or similar agentic development systems seriously Have excellent attention to detail Communicate clearly Can work independently Are proactive when you identify a better solution Enjoy learning new technologies very quickly Strong TypeScript/JavaScript and Python experience is a major plus. Experience with cloud GPU infrastructure, AI inference infrastructure, queues, storage/CDNs and production AI systems is also highly valuable. What we're NOT looking for Please don't apply if your experience is primarily: Creating basic ChatGPT wrappers Building simple no-code AI automations Calling one AI API from a frontend Prompt engineering without software-engineering experience Using Claude Code without understanding the code being generated Using ComfyUI only through downloaded workflows without understanding how they work We need someone capable of deep troubleshooting and technical ownership. Engagement This is intended to become a long-term collaboration. Approximately 30–40 hours per week Minimum expected duration: 6 months Strong possibility of continuing significantly longer We would like someone who can start soon Ideally we will select the right person within the next 1–2 weeks We would prefer someone who can have at least some working-hour overlap with Central European time. How to apply Please start your proposal with the words: GENERATIVE AI This lets us know you've read the complete description. Then tell us: What production SaaS or AI products have you personally developed? Please describe your exact contribution. How are you currently using Claude Code in your development workflow? Be specific. Which image/video generation platforms have you personally integrated through APIs? Please distinguish between APIs you actually integrated and tools you have only used through their UI. Describe your experience with ComfyUI and LoRA training. If possible, include examples of workflows, systems or results you've created. Tell us about one particularly difficult production bug or AI integration problem you diagnosed and solved. Please also include your weekly availability, timezone and monthly rate. We are specifically looking for someone interested in becoming a core long-term technical contributor rather than completing a small isolated task.