Description
AI Software Engineer (AI Agent Development)
Title: AI Software Engineer (AI Agent Development)
Location: Cartersville, GA (Qcells Cartersville Factory)
Department: Smart Intelligence Team
The AI Software Engineer will design, build, and operate AI agent systems that support engineers and managers across Qcells’ solar manufacturing lines (ingot, wafer, cell, and module) at the Cartersville (CTV) and Dalton (DLT) plants. The team runs a production multi-agent platform in which domain agents answer process questions in natural language, query manufacturing data through tools, monitor equipment in real time, send early-warning alerts, and generate reports for engineering and management. This role focuses on agent development: tool design, data integration, LLM orchestration, evaluation, and reliable 24/7 operation on the factory floor. Machine learning models built on vision and process data are developed and served as tools that the agents use.
Responsibilities
- Design and develop LLM-based AI agents for manufacturing domains (ingot, wafer, cell, module), including tool/function calling, multi-step reasoning, intent routing, and conversation context
- Build agent tools that safely access manufacturing data (MES, Oracle/MariaDB process databases, REST data gateways, inspection image stores), with SQL guardrails, query validation, and access controls
- Develop real-time monitoring and alerting agents that detect equipment and process anomalies, attach evidence (charts, images, video), and deliver actionable alerts to engineering chat rooms (e.g., Telegram, Microsoft Teams)
- Build automated reporting and analysis workflows in which agents generate daily/weekly reports, root-cause analyses (RCA), and documents or slides from live production data
- Manage LLM serving and routing across cloud APIs (e.g., Claude) and on-premise open-weight models on local GPU servers, including prompt management, prompt caching, and cost/latency optimization
- Build domain knowledge bases and retrieval (RAG) from SOPs, engineering reports, and team conversations
- Develop and deploy ML models that agents use as tools, such as image-based failure prediction, defect classification, and anomaly detection, with automated retraining and performance monitoring
- Establish evaluation and quality controls for agents: answer verification, user feedback loops, regression tests, and guardrails against incorrect or fabricated answers
- Integrate agents with factory systems such as digital twin platforms, web dashboards, and production simulators
- Work directly with process, equipment, and quality engineers to turn field requests into agent features, and own those features in production
Required Qualifications
- Bachelor’s degree in a quantitative discipline (e.g., Computer Science, Statistics, Industrial Engineering, Mathematics, or a related field)
- 2+ years of relevant experience in software engineering, machine learning, or data engineering
- Hands-on experience building applications on LLM APIs, including tool/function calling, structured outputs, and prompt design
- Strong proficiency in Python, including backend services, asynchronous programming, and REST APIs
- Strong SQL skills and experience working with relational databases
- Experience deploying and operating services on Linux (scheduling, logging, monitoring, troubleshooting)
- Ability to collaborate effectively with colleagues, customers, suppliers, and stakeholders from different technical backgrounds, and to turn ambiguous requests into working tools
Preferred Qualifications
- Professional working proficiency in Korean (speaking, reading, and writing) to collaborate with Korean-speaking engineers and headquarters teams and to build agents that serve Korean-language users
- Experience with agent frameworks and protocols (e.g., Claude Agent SDK, LangGraph, Model Context Protocol)
- Experience with RAG systems, vector databases, and embedding models
- Experience serving open-weight LLMs on premise (e.g., vLLM, Ollama) and managing GPU servers
- Experience with computer vision or time-series machine learning (PyTorch) and MLOps practices
- Experience with manufacturing data (MES, equipment logs, inspection images), especially in solar, semiconductor, or display industries
- Experience with messaging bot integrations (e.g., Telegram, Slack, Microsoft Teams) and web dashboards
- Experience with Docker, Git, and CI/CD
Hanwha Qcells provides equal employment opportunities (EEO) to all employees and applicants for employment without regard to race, color, religion, sex, national origin, age, disability or genetics.