AI Engineer
At a glance
Mid-level AI Engineering role at Fuku. Singapore · full-time.
Pay not stated AI Engineer salary
aimlroles summary, based on the employer's posting.
What you'll do
- Build AI agents for customer service and operational tasks, including workflows that reason, use tools and complete multi-step actions
- Connect agents to internal systems through APIs, automation tools and other integration methods
- Add human review and exception paths for AI actions with higher risk
- Ship LLM applications for classification, summaries, information extraction, decision support and automated responses
- Improve RAG retrieval using customer service knowledge, SOPs, policies and FAQs, then assess answer quality and grounding
What you bring
- You can deliver AI applications into production, rather than focusing only on research or model training
- You can manage AI use cases from identifying an operational problem through ongoing production improvements
- You have experience applying LLMs, agent workflows and RAG to practical customer service needs
Who this fits
This hands-on role suits someone who can turn customer service problems into deployed AI solutions and work with Customer Service, Automation, CRM and Platform teams. The position is onsite in Singapore.
From the employer
About Hytech
Hytech is a leading management consulting firm headquartered in Australia and Singapore, specialising in digital transformation for fintech and financial services organisations. The company delivers end-to-end consulting services and provides robust middle- and back-office solutions that enable clients to optimise operations, enhance efficiency, and stay ahead in a fast-evolving digital landscape. Hytech’s client portfolio includes top global trading platforms and leading crypto exchanges.
With more than 2,000 professionals worldwide, Hytech has a strong and growing international presence, with offices across Australia, Singapore, Malaysia, Taiwan, the Philippines, Thailand, Morocco, Cyprus, Dubai, and beyond.
Role Overview
We are seeking a hands-on AI Engineer to design, build, and deploy practical AI solutions within our Customer Service ecosystem. This role is focused on delivering production-ready AI solutions, not research or purely model-training. The successful candidate will collaborate closely with Customer Service, Automation, CRM, and Platform teams to identify operational pain points and translate them into effective AI solutions. Key responsibilities include building AI Agents, LLM-powered workflows, knowledge/RAG solutions, intelligent decisioning, and AI-driven automation, with integration into platforms such as Zendesk, CRM systems, APIs, and Automation solutions. The ideal candidate will independently manage AI use cases from business problem identification through to continuous improvement in production.
Key Responsibilities
AI Agent & Agentic Workflow Development
- Design and develop AI Agents for Customer Service and operational use cases.
- Build agentic workflows involving reasoning, decision-making, tool calling, and multi-step task execution.
- Enable AI Agents to interact with internal systems through APIs, automation tools, and other integration mechanisms.
- Design human-in-the-loop and exception-handling mechanisms for higher-risk scenarios.
LLM & Applied AI Development
- Develop production-ready applications using commercial and/or open-source LLMs.
- Implement prompt engineering, structured outputs, function/tool calling, context management, and LLM orchestration.
- Design AI solutions for classification, summarization, information extraction, decision support, response generation, and workflow automation.
- Evaluate different models and approaches based on accuracy, latency, cost, security, and business requirements.
Knowledge & RAG
- Design and improve RAG-based knowledge retrieval systems.
- Work with structured and unstructured Customer Service knowledge, SOPs, policies, and FAQs.
- Improve retrieval quality, grounding, citation, chunking, metadata, and knowledge accuracy.
- Develop evaluation framewor