Senior AI Engineers at Treasure AI lead the development of sophisticated AI features — owning LLM integrations, agentic workflows, and AI product experiences from design through production. You own the AI product layer — LLM integration, prompt design, and evaluation — working with Software Engineers to build the features that consume and deliver it. At this level you don’t just execute within established patterns — you define them for your pod (a 3–4 person delivery team) and elevate everyone around you. Success means shipping AI features that work reliably at scale, setting the LLM integration standard for your pod, and staying on the leading edge of what’s possible with language models.
Treasure AI moves marketing and data teams beyond legacy martech SaaS toward autonomous, ROI-driven AI agents that operate continuously, while keeping the human in the loop. With built-in governance, Treasure AI is powered by an always-up-to-date understanding of each customer for AI to act on.
The result is always-on marketing execution that compounds over time: smarter engagement, stronger retention, and measurable growth.
Treasure AI’s agentic experience platform is available across web, mobile, and desktop — bringing your customer intelligence to every workflow, wherever your team works.
Responsibilities
- Lead the design and implementation of AI-powered features: LLM integrations, retrieval-augmented generation, agentic pipelines, and AI-augmented workflows
- Own prompt engineering rigor — structure prompts systematically, evaluate outputs, iterate based on production signal, and document what works
- Develop LLM integration patterns for the pod: streaming responses, function calling, context window management, fallback handling, and evaluation
- Understand the AI solution space well enough to recommend the right approach — knowing when to reach for prompting, RAG, fine-tuning, or agentic patterns, and when to bring in additional expertise
- Contribute production-grade code in Ruby/Rails and TypeScript/React; hold a high bar for code quality in AI and non-AI features alike
- Use Claude Code and GitHub Copilot fluently; critically review AI-generated code and help teammates do the same
- Work with our customer-facing agent platform to build, iterate, and deploy AI capabilities
- Independently scope AI feature work within the team’s 3-week delivery cadence, accounting for experimentation cycles and evaluation needs
- Own on-call shifts; develop operational instincts for AI systems in production — including latency, cost at scale, multi-tenant routing, and customer data boundaries
- Ship AI features iteratively — deliver a working version early and refine based on real usage
- Conduct deep code reviews on AI feature work; help teammates reason about model behavior, prompt design, and integration robustness
- Share LLM integration knowledge across the pod and contribute to cross-pod AI engineering discussions
- Help peers grow their AI engineering skills through pairing, documentation, and structured feedback
- Champion shared ownership of AI systems — avoid knowledge silos around model behavior and integration details
Requirements
- 4–7 years of software engineering experience with significant focus on AI/ML features or LLM-powered product development
- Strong proficiency in prompt engineering, LLM API integration, and building AI features that perform reliably in production
- Experience with RAG, agentic workflows, or function-calling/tool-use patterns with language models
- Experience making AI solution tradeoffs in production — choosing between prompting, RAG, fine-tuning, and agentic approaches based on real constraints
- Solid full-stack skills in Ruby/Rails and TypeScript/React
- Fluency with Claude Code, GitHub Copilot, and our customer-facing agent platform
- Experience evaluating LLM outputs systematically — evals, test cases, production monitoring
- Operational awareness of AI systems in production: latency management, cost at scale, multi-tenant routing, and customer data boundaries
- Track record of meaningful code reviews that grow teammates’ technical capabilities
Compensation
¥10,000,000 ~ ¥14,000,000 annually.
