PyTorch Jobs

Job Introduction

We are building the most advanced AI Agent for the Web3 industry, leveraging the largest proprietary dataset in the field. We seek a core algorithm engineer to architect AI Agent systems, optimize end-to-end RAG pipelines, implement LLM training/alignment, and deploy scalable.

Core Responsibilities

1. Develop AI Agent Systems: Build intelligent search and task execution agents using ReAct, planning, and multi-agent frameworks (e.g., LangGraph, Dify, CrewAI)

2. Optimize End-to-End RAG Pipelines: Build and refine efficient RAG systems from ingestion, chunking, and embedding to hybrid vector search (OpenSearch), implementing precise grounding and citation

3. LLM Training & Alignment: Conduct advanced post-training (SFT, RLHF, continual pretraining) and align models for reliable JSON-schema function calling and external tool usage

4. Automated Evaluation & Iteration: Build offline/online evaluation pipelines using synthetic QA, retrieval metrics, and hallucination detection to continuously improve system performance and stability

Qualifications

1. Bachelor's or Master's degree in Computer Science, AI, Machine Learning, or a related field

2. 3+ years of experience developing AI systems, with a focus on RAG, Agent architectures, or LLM training/optimization

3. Proficiency in Python and key ML frameworks (PyTorch/TensorFlow), with experience in distributed training and high-performance inference

4. Hands-on, in-depth experience in at least two of the following domains:

   • End-to-end RAG pipeline development and optimization with OpenSearch/vector databases

   • AI Agent framework development (LangGraph, CrewAI, ReAct)

   • Advanced LLM training (SFT, RLHF, LoRA) and alignment techniques

5. Excellent problem-solving and systems thinking skills. Passion for Web3 and AI is a plus

Key Outcomes

• Deliver a high-accuracy, low-latency AI Agent system to power intelligent Web3 applications

• Achieve continuous improvement in RAG retrieval accuracy and establish an automated evaluation and iteration loop

• Drive LLM performance optimization