Generative AI Engineer (Python)

SwipeyAI is a fast-growing 18+ AI companion platform combining conversational AI, generative media, gamification, subscriptions, and creator monetization. 💡 What you’ll do: • Build and improve the AI pipelines powering SwipeyAI companions. • Own parts of the chat pipeline: prompt construction, personality conditioning, memory, context, and reply post-processing. • Integrate and evaluate new LLM, image, and video generation models. • Train and retrain LoRAs for character consistency across image and video generation. • Build multi-step and agentic generation workflows with automated validation. • Run structured prompt experiments and model evaluations. • Work with GPU inference infrastructure and optimize quality, latency, and cost. • Monitor model performance, generation failures, and production metrics. 🛠 Tech / tools: Python, LLMs, diffusion models, LoRA, SDXL-class models, image-to-video, Langfuse, Docker, AWS, MongoDB, RunPod / fal.ai or similar GPU providers. 🎯 What we expect: • 3+ years of professional Python development. • Strong production backend experience: async patterns, queues, microservices. • Hands-on experience integrating LLMs or generative models into a real product. • Practical experience with diffusion pipelines: LoRA training / fine-tuning, model evaluation, image or video generation. • Experience building multi-step generation or agentic workflows. • Confidence working with Docker, CI, monitoring, GPU inference, and production systems. • Good English. • Comfortable working with an adult-content (18+) product. ➕ Nice to have: • Langfuse or similar prompt/eval tooling. • ComfyUI, Wan or other video generation models. • vLLM / TGI or other self-hosted serving. • Conversational memory systems. • Content moderation / safety-classifier experience.