AI-Native Full-Stack Engineer

About This Role

• Join our R&D squad as an AI-Native Full-Stack Engineer. Forget legacy maintenance - your mission is to rapidly transform leadership visions into high-quality Proof of Concepts (PoCs).
• As a "Technical Scout," you will leverage AI agents to automate heavy lifting, validate features, and deliver standardized, production-ready hand-offs for the main engineering team. Solid fundamentals are required; mastery of AI orchestration is essential.

Job Description

• Rapid Prototyping with Agents: Utilize Agentic Coding workflows to quickly build MVPs. Verify the feasibility of integrating third-party services or new technologies and turn abstract requirements into interactive applications in record time.
• Research & Knowledge Handover: Use AI to accelerate technical documentation reading and solution comparison. You are responsible for the "0 to 1" exploration and must write comprehensive developer documentation (including Context setup) to ensure the "1 to 100" main team can smoothly integrate your work.
• AI Workflow Architecture: You won’t just write code; you will design the flows that allow AI to write code for you. You will build and maintain Prompt Templates and Agent Toolchains suitable for R&D tasks.

Requirement

1. Foundation & Experience

• 2+ years of full-stack development experience.
• Strong independent judgment regarding code quality and architecture. You must have the expertise to correct AI errors and validate generated solutions, rather than being misled by them.

2. Agentic Workflow & Tool Proficiency

• You are proficient with coding agents like Claude Code or OpenCode. You are comfortable working in a terminal environment where AI executes commands, edits files, and manages git operations autonomously under your supervision.
• You understand the concepts of REPL (Read-Eval-Print Loop) in AI coding and know how to guide an agent through complex refactoring tasks step-by-step.

3. Context Management & RAG Application

• Understanding of Context: You understand how AI tools retrieve or lose information. You can debug generic AI responses by manually curating context (managing what files or docs fit in the context window) to ensure tools reference the correct "source of truth."
• Prototyping RAG Features: You don't need to build Vector DB algorithms from scratch, but you must know how to implement basic RAG (Retrieval-Augmented Generation) pipelines to build feature prototypes (e.g., connecting a vector database to a chatbot demo).

4. Tool Integration & Agent Skills

• Agent Skills Utilization: Beyond simple MCP connections or API calls, you are adept at selecting and utilizing the right "Agent Skills" within tools like Claude Code to accelerate development.
• Toolchain Mastery: You possess your own AI development toolchain. You know how to grant and manage permissions for Agents to execute file I/O or run scripts safely.

5. Rigorous Quality Control

• You never blindly trust AI output. You have strict standards for Code Review and testing.
• You integrate Linting, Type Checks, and automated testing into the AI generation loop. You ensure that the code produced by Claude Code/OpenCode is not just "functional" but robust, clean, and compliant with team standards before handover.

Nice-to-Have (Bonus Qualifications)

1. Specification-Driven Development (SDD) & Precision AI Coding

• Methodology & Research: Experience with Specification-Driven Development (SDD) or similar methodologies aimed at making AI coding more precise. Whether through communities like sdd.tw or other advanced research/courses, you are committed to finding ways to bridge the gap between human specs and machine code.
• Spec-to-Code Proficiency: You excel at transforming requirements into Testable Code (executable specifications/tests) before implementation. You use these tests as "guardrails" to ensure the AI's output strictly adheres to business logic.

2. Advanced Multi-Agent Architecture

• Experience designing Multi-Agent Systems where different AI roles (e.g., Architect Agent, QA Agent, Coder Agent) collaborate to solve complex problems.
• Ability to build custom MCP Servers or develop proprietary Agent Skills/Tools.

3. Resource Efficiency & Security Awareness

• Cost Management: You monitor Token usage during development (especially when using high-cost models) and know when to switch to smaller models for simple tasks to optimize ROI.
• Security: You have strong data classification awareness. You know how to prevent sensitive data (API Keys, PII) from being accidentally read by AI agents and understand how to defend against Prompt Injection attacks in your prototypes.

• CV Screening
• HR Contact
• Hiring Manager Interview
• Take-home Assignment (Optional)
• Team Interviews (Optional)
• Offer!