AI Engineer

Red Begonia Limited; Hong Kong; On-site; HKD 1,200,000 - 1,800,000; Full time

Job summary

RedotPay is a global crypto payment fintech building AI-powered capabilities for digital finance. We are expanding the AI Agent Runtime with an architecture that orchestrates agents, tools, and memory across multi-turn interactions.

The role focuses on enterprise-grade agent platforms, runtime performance, and integrating evolving AI technologies into production workflows to enable secure and scalable payments.

Qualifications

• 5+ years of software engineering experience, with AI/Agent R&D exposure.
• Bachelor’s degree or higher in Computer Science or related field.
• Strong knowledge of agent design patterns and LLM integration.
• Experience with backend stacks and real-time communication (e.g., WebSockets).
• Familiarity with streaming protocols and AI workflows.

Responsibilities

• Architect and build core modules for the AI Agent Runtime, including orchestration, Tool Calling, memory management, and context orchestration.
• Drive enterprise-grade Agent capabilities, including plugin architecture and multi-LLM vendor adaptation.
• Work on runtime stability and performance, including long-context management, concurrency, token optimization, and caching.
• Track cutting-edge AI technologies (Function Calling, Multi-Agent systems, MCP) and translate into production features.
• Collaborate on AI-native payment workflows and end-to-end execution scenarios.

Job description

RedotPay is a global crypto payment fintech that integrates blockchain solutions into traditional banking and finance infrastructure. Our user-friendly crypto platform empowers millions worldwide to spend and send crypto assets, providing faster, more accessible and inclusive financial services. We are dedicated to advancing financial

inclusion for the unbanked and supporting crypto enthusiasts, driving the global adoption of secure and flexible crypto-powered financial solutions. Join us in shaping the future of finance and making a meaningful impact on a global scale.

Key Responsibilities

• Agent Runtime Core Architecture & Development: Architect and build core modules for the AI Agent Runtime, including the Agent orchestration engine, Tool Calling pipeline, multi-turn conversation memory management, and context orchestration, establishing a stable and efficient execution environment.
• Enterprise Agent Platform Expansion: Drive the delivery and scaling of enterprise-grade Agent capabilities, including plugin architecture, multi-LLM vendor adaptation, multimodal integration (text/images/files), as well as Agent orchestration and workflow engines.
• Runtime Stability & Performance Optimization: Responsible for the stability and performance tuning of the Agent Runtime—managing long-context windows, concurrent scheduling, Token consumption optimization, rate limiting, caching, and fallback strategies to ensure high availability in production environments.
• Cutting-Edge AI Tech Tracking: Continuously monitor evolving LLM and Agent technologies (e.g., Function Calling, Structured Outputs, Multi-Agent systems, Model Context Protocol / MCP), rapidly translating front-tier technical advancements into production-ready features.
• AI Payment Protocol Exploration: Collaborate in exploring AI-native payment workflows based on emerging AI Payment protocols (e.g., A2A, AP2, X402), driving end-to-end execution scenarios: User → AI Agent → Payment Capability Calling → Automated Transaction Execution

Qualifications & Requirements

• Experience & Education: Bachelor’s degree or above in Computer Science or related fields; 5+ years of software engineering experience, with at least 2+ years of hands-on experience in AI Agent system R&D.
• Agent Design Patterns: Deep mastery of mainstream Agent design patterns with the ability to choose and combine optimal solutions based on business scenarios (e.g., Single Agent, ReAct, Plan-and-Execute, Reflection, Tool Use / Function Calling, Multi-Agent Collaboration). Strong understanding of each pattern's application boundaries, Token costs, stability, and interpretability trade-offs.
• LLM Application Engineering: In-depth knowledge of Agent architecture core design principles and tuning strategies, including Prompt Engineering, Function/Tool Calling, short/long-term memory, Chain of Thought (CoT), complex multi-turn reasoning, SOP automation, and autonomous decision-making.
• Backend Stack & Real-Time Communication: Proficient in at least one modern backend stack (Spring Boot, FastAPI, or Node.js). Able to independently design Agent service APIs and develop modules across databases (MySQL/PostgreSQL/Redis/Vector DBs), message queues, and asynchronous tasks. Strong understanding of streaming protocols and AI communication patterns (e.g., SSE, WebSockets).
• AI-First Engineering Mindset: Daily heavy user of AI coding tools (e.g., Claude Code, Cursor, Codex), integrating LLMs into the core workflow rather than using them merely as passive assistants. Views Tokens as production inputs—knowing how to make precise engineering trade-offs between output quality, cost, and latency, with a strong ability to drive Agent iteration through Evals.