AI Engineer

AI Product Engineer (Mid to Senior)

📍 Tokyo | Office-first Hybrid (4 days office / 1 day remote)

Our client is searching for a talented AI Product Engineer to join a fast-growing AI SaaS company building cutting-edge creative technology for global enterprise brands. This is a unique combined role sitting at the intersection of practical ML evaluation and production backend engineering — bridging the gap between AI model capabilities and commercial-grade creative workflows.

SUMMARY:

• Combined ML evaluation and backend integration role responsible for benchmarking and selecting AI models, building production pipelines, and owning the full quality data flywheel that drives continuous product improvement
• High-ownership, end-to-end delivery position translating ambiguous product feedback on visual quality into robust, shippable features — from model evaluation and pipeline tuning through to production backend integration and observability
• Tokyo office-first hybrid role (4 days in office); fluent English required, Japanese is a plus; open to mid and senior level candidates with 4+ years of backend or applied ML engineering experience

RESPONSIBILITIES:

• Benchmark, evaluate, and select optimal AI models and providers for visual feature workflows including upscaling, retouching, and lighting control, using structured evaluation sets and benchmarking scripts
• Partner with backend engineers to design robust APIs, async job workers, retries, and error-handling systems for media generation pipelines
• Build data capture mechanisms to establish evaluation sets and continuously drive model routing and product quality decisions
• Translate ambiguous creative feedback into concrete hypotheses, testable benchmarks, and production-ready code
• Own end-to-end delivery across the AI feature lifecycle — from initial evaluation through production integration, observability, and continuous improvement

REQUIREMENTS:

• 4+ years of professional backend or ML engineering experience with strong Python engineering practices
• Hands-on experience integrating AI models, diffusion pipelines, or multimodal APIs into live production applications
• Solid backend fundamentals including API design, async job processing, queue management, status tracking, and error handling
• Ability to quantitatively and qualitatively diagnose visual output failures and translate findings into actionable product improvements
• Fluent English required; Japanese is a plus

Nice to Have:

• Experience with image or video generation, editing, or upscaling providers such as diffusion models, transformers, CLIP/VLMs, super-resolution, or multimodal evaluation tools
• Experience building eval datasets, benchmark scripts, annotation workflows, A/B comparisons, or experiment tracking systems
• Background in ecommerce, beauty, fashion, photo/video apps, adtech creative tooling, or asset-heavy production workflows