AI Application Engineer (Agentic AI)

About the Company

SSMC (Systems on Silicon Manufacturing Company Pte. Ltd.) is a semiconductor manufacturing joint venture formed by NXP and TSMC. The company provides efficient and cost-conscious fabrication services through a fully equipped SMIF cleanroom setup, complete equipment automation, and established wafer production processes. It is seeking motivated, creative, and capable professionals to join its team.

Role Overview

This position sits within the Fab Automation & Application Department and focuses on supporting AI-powered applications used across the fab environment. The role involves keeping production AI systems running smoothly, improving model performance, and helping deploy new solutions across the facility.

Key Duties

• Oversee and support AI applications that are already live in production.
• Watch model performance indicators and spot any drop in accuracy or reliability.
• Carry out retraining based on defined processes and operational playbooks.
• Keep records of performance trends, maintenance work, and model changes.
• Roll out newly introduced AI applications across the fab environment.
• Work with base models and gather equipment-specific data for fine-tuning.
• Apply model upgrades through prebuilt pipelines and verify the results.
• Check model performance again after updates or equipment integration.
• Detect image-quality problems, data drift, and other model-related anomalies.
• Partner with users to fix application or hardware issues that affect model output quality.
• Handle basic troubleshooting for AI pipeline components.
• Escalate difficult technical issues and coordinate with vendors when needed.
• Build and carry out improvement actions to strengthen accuracy and response quality.
• Work alongside the IT team to keep training and validation datasets clean and well organized.
• Prepare and maintain technical documents for support and maintenance activities.
• Produce routine reports on model health and overall system status.
• Ensure work follows data governance and quality requirements.

Candidate Profile

The ideal applicant should hold a bachelor’s degree in an engineering, computer science, data science, or similar technical discipline. The role calls for 3 to 5 years of hands-on experience with machine learning models, large language model concepts, and prompt engineering used in solution delivery. Exposure to AI platforms, NLP tools, and frameworks such as TensorFlow or PyTorch is expected. Experience with lighthouse adoption or deployment is also relevant.

Applicants should have familiarity with data engineering, large-scale data handling, data modeling, and warehousing. Strong programming ability in Python, C++, JavaScript, and R is required, along with practical use of visualization tools such as Power BI, Tableau, matplotlib, or seaborn, plus web application development experience. Knowledge of MLOps tools like MLflow or Kubeflow is important, as is comfort with pandas, numpy, SQL databases such as PostgreSQL, and cloud platforms including AWS, Azure, or Google Cloud, as well as GitHub. Basic understanding of Docker and Git is needed.

In addition, the role requires someone who is detail-focused, well organized, able to work independently, and comfortable following standard operating procedures. A willingness to learn quickly and adapt to new tools, workflows, and technologies is essential.

Equal Employment Opportunity

SSMC is committed to fair employment practices and equal opportunity for all candidates. Hiring decisions are based on merit and made without discrimination on the basis of age, race, gender, religion, marital status, family responsibilities, disability, or any other protected attribute under applicable laws. The company also states compliance with the Tripartite Guidelines on Fair Employment Practices (TGFEP), the Singapore Food Safety and Security Act 2025 (FSSA 2025), and the Singapore Code of Advertising Practice.

Additional Information

Work location: Pasir Ris, East Region, Singapore. Employment type: full-time, onsite.