Data Scientist

Description

About the Role -

We’re seeking a Data Scientist for Agentic AI to help build AI agents that drive yield improvement, root-cause discovery, and process optimization for leading semiconductor companies.

You’ll combine modern data science and machine learning with LLM- and agent-based techniques to power the next generation of agentic data systems in semiconductor manufacturing and product engineering. Your work will turn messy, high-dimensional data into agents that can continuously monitor, diagnose, and recommend actions across the semiconductor lifecycle.

You will work closely with process, product, and test engineers, as well as data engineers and platform developers, to ingest complex data across the semiconductor lifecycle and build analytical pipelines, models, and agents that surface insights and take action automatically.

Direct semiconductor yield and process experience is not required, but familiarity with yield analysis and semiconductor manufacturing concepts is a strong plus.

Key Responsibilities -

• Data Integration & Management
• Ingest and unify data from diverse sources across the semiconductor lifecycle (process, test, assembly, and field / system-level data).
• Handle structured data (SQL, CSV, metrology logs, time-series) and unstructured data (reports, images, logs).
• Develop pipelines to clean, align, and normalize large datasets across manufacturing stages and product lines.

• Analytics, Modeling & Agent Intelligence
• Design and implement algorithms for excursion and anomaly detection, trend monitoring, and yield / quality loss pattern recognition.
• Perform advanced correlation analysis across process parameters, test metrics, and design variables.
• Support root-cause analysis by developing interpretable statistical and machine learning models to identify likely drivers of excursions and quality drift.
• Develop feature extraction and dimensionality-reduction methods suitable for high-dimensional industrial and manufacturing data.
• Collaborate with ML and agent engineers to design agents that can call tools, query data, run analyses, and iteratively refine their own hypotheses.

• Visualization & Insight Delivery
• Create interactive dashboards and visualizations (e.g., wafer maps, trend charts, pareto analyses) to communicate findings to process and product engineers.

• Research, Innovation & Agentic AI
• Evaluate and implement modern AI/ML approaches for yield optimization, root-cause analysis, and reliability monitoring (e.g., graph-based methods, representation learning, LLM-assisted analysis).

M.S. or Ph.D. in Computer Science, Electrical Engineering, Applied Physics, Statistics, Applied Mathematics, or a related quantitative field (or equivalent practical experience).

• 4+ years of experience in applied data science, machine learning, or advanced analytics.
• Experience working with large, complex, multi-source datasets (e.g., time-series, sensor, manufacturing, or hardware data).
• Proficiency in Python (e.g., pandas, scikit-learn, Num Py) and common ML / data science workflows.
• Strong experience with databases and querying (SQL and No
  
  SQL).
• Experience with data visualization tools and frameworks (e.g., Plotly or similar).
• Strong grasp of statistical methods such as outlier detection, correlation analysis, regression, clustering, causal inference, and time-series analysis.

• Comfortable working with both engineering and data teams, as well as domain experts.
• Strong analytical thinking, curiosity, and ability to explain complex findings clearly to non-specialists.
• Practical mindset — focused on building usable analytics and agents that deliver real operational impact.

• Familiarity with semiconductor yield analysis and manufacturing processes (e.g., yield metrics, bin analysis, wafer maps, process flows).
• Experience with wafer-level and bin-level data analysis, wafer map pattern recognition.
• Experience using or developing yield management systems (YMS) or manufacturing data platforms.
• Experience with multi-modal or foundation models (LLMs, VLMs) applied to industrial or manufacturing data.
• Exposure to LLM-assisted data analysis, AI-based…