What you will do
- Build and deploy LLM-based systems for internal use (Agentic BI, chatbots, automation tools);
- Train, fine-tune, and evaluate LLMs (SFT, RAG optimization, preference tuning);
- Design and maintain large-scale data pipelines (ETL, feature pipelines);
- Develop and scale AI Agents (tool usage, planning, memory, multi-step reasoning);
- Optimize system performance, latency, and cost at scale;
- Build internal tools (NLP → SQL, NLP → dashboards) for Data/Business teams.
What you will need
- 4+ years of working experience;
- Hands-on experience with LLMs:
- Fine-tuning (LoRA/QLoRA or full fine-tuning);
- Evaluation (offline + A/B testing);
- Strong understanding of:
- RAG, embeddings, vector search;
- Prompt engineering & context optimization;
- Experience building AI Agents (LangChain, LangGraph or custom orchestration);
- Experience with data systems:
- Apache Spark;
- Apache Doris;
- Strong Python + backend (API, async, microservices);
- Good system design thinking (quality – latency – cost trade-offs).
Nice to Have
- Experience training/adapting open-source LLMs (LLaMA, Mistral);
- Familiar with vector DBs (Milvus, FAISS, Pinecone);
- Knowledge of distributed systems (Kafka, Redis);
- Experience with inference optimization (vLLM, TensorRT-LLM).
Bonus (Strong Advantage)
- Frontend experience (React, Next.js, or similar) to build internal AI tools/UI
(e.g., chatbot UI, Agent dashboards, BI interfaces); - Built production-grade AI Agents / LLM systems;
- Open-source projects or strong GitHub portfolio;
- Strong product thinking & business understanding.
