Job description for Backend Ai Engineer at PT Unilabel Pakindo Jaya
We are looking for an experienced Backend AI Engineer to join our team. In this role, you will focus on designing, building, and maintaining scalable backend architectures using Java or Golang, while integrating cutting-edge Artificial Intelligence (AI) and Large Language Models (LLMs) into our core business systems. You will be responsible for the entire AI integration lifecycle, from LLM workflow orchestration and advanced RAG optimization to implementing secure guardrails.
Key Responsibilities:
Backend Development: Design, build, and maintain robust, scalable, and highly efficient backend services (APIs, Microservices) using Java or Golang.
AI & LLM Integration: Directly integrate AI models (particularly Claude and other leading LLMs) into production-level business applications.
RAG Architecture & Optimization: Build and optimize end-to-end Retrieval-Augmented Generation (RAG) pipelines, managing everything from data ingestion to retrieval.
Workflow Orchestration: Design and automate complex LLM workflows using n8n or similar workflow automation frameworks.
Agentic Systems & Guardrails: Develop autonomous AI Agent architectures and implement strict Guardrails to ensure all AI inputs and outputs are safe, ethical, reliable, and within defined system boundaries.
AI Observability & Evaluation: Monitor, evaluate, and continuously improve AI application performance and latency using observability platforms such as LangSmith, Arize, or Braintrust.
Qualifications & Requirements:
Software Engineering Background: Proven professional experience in backend software development using Java or Golang.
AI Experience (Minimum 2 Years): At least 2 years of direct experience developing AI-based applications or being deeply involved in AI projects that have been successfully integrated into real-world business systems.
RAG Mastery (Mandatory): Deep understanding and hands-on experience with advanced RAG architectures, specifically encompassing:
Text/Document Chunking strategies.
Generation and management of Embeddings.
Implementation of Hybrid Search (combining vector search with lexical/keyword search).
Reranking mechanisms (e.g., Cross-Encoders, Cohere Rerank) to improve retrieval accuracy.
LLM Ecosystem & Tools: Strong hands-on experience with the following technologies:
Commercial LLM APIs (strong focus on Claude).
Workflow automation tools like n8n.
AI evaluation and tracing platforms: LangSmith, Arize, or Braintrust.
Advanced AI Concepts: Solid understanding of building AI Agents and implementing safety mechanisms / Guardrails for LLM interactions.
Nice to Have (Bonus):
Hands-on experience working with Vector Databases (e.g., Pinecone, Qdrant, Milvus, Weaviate, or pgvector).
Advanced prompt engineering skills for optimizing model responses.
Familiarity with Event-Driven architectures or Message Brokers (Kafka, RabbitMQ) for asynchronous AI task processing.
