Job Description
Full details about the role and requirements
Yukerja Summary
The Junior AI Engineer role at WIRATEK SOLUSI ASIA is curated from Glints (category Teknologi & IT). Note the work location (Purwokerto Barat) before applying. Yukerja.com is not the employer — applications are handled on the official source site.
We build production AI systems for enterprise clients across many industries in Indonesia, running our own GPU inference infrastructure end to end. We're hiring an AI Engineer to design and build AI architecture (RAG pipelines, agentic systems, and the serving stack behind them) and ship it to production.
What You'll Do
- Build RAG pipelines end to end: ingestion, chunking, embedding, indexing, hybrid search (dense + sparse, RRF), reranking, and generation.
- Design agentic workflows with tool calling, routing, and state management, including connectors via MCP.
- Deploy and tune self-hosted LLMs (vLLM) on NVIDIA GPUs; optimize VRAM, batching, concurrency, and caching.
- Ship AI features into backend services (FastAPI, Celery, NestJS, Socket.IO / Redis) with ACL, audit logging, and evaluation.
- Work with PMs, backend, and frontend developers, and document in Bahasa Indonesia.
Required
- Real LLM projects you can show and walk us through (work, internships, open-source, or personal).
- Strong Python (async); solid grasp of RAG, prompting, tool use, and agentic patterns.
- Hands-on with vector databases, embedding models, FastAPI, and Celery.
- Some exposure to GPUs and model serving.
Nice to Have
- vLLM or Ollama and GPU optimization; Milvus and hybrid search / reranking.
- MCP, PydanticAI, LangChain, LangGraph; NestJS / Node, Socket.IO, Redis.
- Domain experience in enterprise knowledge platforms, medical/clinical AI, Human Capital / HR analytics, or AI CV / resume screening.
Stack
PydanticAI · LangChain · LangGraph · FastAPI · Celery · Milvus · vLLM · Ollama · PostgreSQL · NestJS · Socket.IO · Redis