AI Engineer

2 months ago

Ōsaka, JapanMid Level / Senior

Responsibilities

Architect multi-agent systems with branching logic, exception handling, tool integrations, memory, context management, state persistence, guardrails, and human-in-the-loop escalation.
Own end-to-end LLM adaptation strategy, including fine-tuning versus RAG versus prompt engineering, dataset preparation, benchmarking, model versioning, GPU training, model merging, quantization, and production serving.
Build RPA task bots as execution layers within agentic workflows and architect agent-to-RPA handoff and exception-management logic.
Build agent evaluation frameworks and implement observability and tracing for production systems.
Implement CI/CD for agent and model deployments and diagnose hallucinations, tool misuse, and runaway costs.
Lead use-case discovery workshops, independently lead solution architecture, communicate architecture trade-offs to nontechnical stakeholders, and mentor junior engineers.

Bachelor’s or Master’s degree in Computer Science, AI/ML, Data Science, or equivalent practical experience.
4–7 years of software engineering experience and 3+ years building production-grade automation solutions.
Demonstrated end-to-end delivery of agentic AI systems or complex enterprise RPA prototypes.
Production-quality Python engineering experience.
Experience with agentic AI frameworks, multi-agent patterns, LLM fine-tuning and adaptation, RAG and vector infrastructure, model serving, and enterprise RPA.
Experience with LoRA/QLoRA, HuggingFace PEFT or Unsloth, dataset preparation, evaluation benchmarking, model versioning, vLLM, GPTQ, and GGUF.
Experience with Pinecone, Weaviate, or Qdrant and embeddings and retrieval evaluation.
Experience with UiPath, Automation Anywhere, or Power Automate in production.
Experience with cloud AI services such as Bedrock, Azure, OpenAI, or Vertex AI.
AWS, Azure, GCP, UiPath, or Automation Anywhere certifications are preferred.

1,001-5,000 employees