Junior AI Engineer (LLM/ Generative AI)

YCH Group Pte Ltd; Singapore; On-site; SGD 60,000 - 120,000; Full time

Job summary

YCH Group Pte Ltd is seeking a developer to build and maintain LLM-powered applications, including chatbots and document processing pipelines. You will design RAG pipelines, optimize embedding strategies, and manage context windows to improve retrieval accuracy.

You will refine prompt libraries, versions, and evaluation metrics, and collaborate with analysts to deliver AI features that integrate with existing Java APIs.

Qualifications

• 1+ years working or project experience on LLM or AI application development.
• Experience building dynamic UIs with React and Next.js for AI streaming responses.
• Design scalable backend API routes and queues with Node.js and TypeScript.
• Hands-on experience with LLM frameworks (LangChain, LangGraph, LlamaIndex).
• Understanding of RAG architecture, vector databases, and embeddings.
• Experience consuming LLM APIs (OpenAI, Azure OpenAI, Vercel AI, Anthropic Claude).
• Solid software engineering fundamentals: version control, testing, documentation.
• Java development background – beneficial for logistics contexts.
• Fine-tuning experience with open-weight models (Llama, Mistral, Phi).
• Knowledge of logistics domain is advantageous.
• Experience with agentic AI frameworks (AutoGen, CrewAI, LangGraph).

Responsibilities

• Build, iterate, and maintain LLM-powered applications (chatbots, document processing, analytics, search).
• Design and optimise RAG pipelines: chunking, embeddings, retrieval tuning, context windows.
• Develop and refine prompt engineering frameworks, libraries, versioning, and evaluation.
• Fine-tune open-weight models on domain data to improve accuracy and reduce costs.
• Integrate LLM capabilities into legacy Java APIs with clean abstraction layers.
• Implement automated evaluation pipelines for faithfulness, relevance, and hallucinations.
• Collaborate with Business Analysts to translate use cases into production AI features.
• Contribute to internal AI documentation, prompt libraries, and reusable components.

Skills

UI Engineering

AI Backend Orchestration

LLM Frameworks

RAG Architecture

Embeddings

OpenAI APIs

Version Control

Testing

Documentation

Java

Fine-tuning Open-weight Models

Logistics Domain Knowledge

Agentic AI Frameworks

Tools

LangChain

LangGraph

LlamaIndex

OpenAI / Azure OpenAI

Job description

• Build, iterate, andmaintain LLM-powered applications — including chatbots, document processingpipelines, predictive analytics interfaces, and intelligent search systems
• Design and optimise RAG(Retrieval Augmented Generation) pipelines: chunking strategies, embeddingmodel selection, retrieval tuning, and context window management
• Develop and refine promptengineering frameworks — maintaining prompt libraries, evaluating promptperformance, and implementing prompt versioning
• Fine-tune open-weightmodels (Llama, Mistral) on YCH-specific logistics data to improve domainaccuracy and reduce inference costs
• Integrate LLM capabilitiesinto YCH's legacy Java application APIs — building clean abstraction layersthat allow AI features without full system rewrites
• Implement LLM evaluationpipelines using automated scoring (faithfulness, relevance, hallucination rate)and human evaluation frameworks
• Collaborate with BusinessAnalysts to translate new use case specifications into production AI features
• Contribute to internal AIdocumentation, prompt libraries, and reusable component libraries

Job Requirements:

• 1+ years working or projectexperience on LLM or AI application development
• User InterfaceEngineering: Build dynamic, real-time UI components using React and Next.js tohandle AI streaming responses, multi-step agent status indicators, and markdownformatting.
• AI BackendOrchestration: Design scalable backend API routes and background event queuesusing Node.js and TypeScript.
• Hands-on experience withLLM frameworks: LangChain, LangGraph, LlamaIndex, or equivalent
• Understanding of RAGarchitecture, vector databases, and embedding models
• Experience consuming LLMAPIs (OpenAI, Azure OpenAI, Vercel AI, Anthropic Claude)
• Solid software engineeringfundamentals: version control, testing, code review, documentation
• Java development background— ideal for reskilled internal candidates with logistics context
• Fine-tuning experience withopen-weight models (Llama, Mistral, Phi)
• Knowledge of logisticsdomain: freight documents, customs, route planning is favoured.
• Experience with agentic AIframeworks (AutoGen, CrewAI, LangGraph)