The challenge: AI Agents for Scientific Workflows 

As artificial intelligence (AI) becomes increasingly integrated into everyday life, scientists are using advanced frontier models to accelerate scientific discovery. This year, students will design agentic AI pipelines built on foundational vision and language models to improve workflows for materials discovery and additive manufacturing. 

Participants will create task-specific agents capable of performing data exploration, visualization, computer vision tasks, and analytical workflows. These agents will then be coordinated to generate meaningful insights from real-world materials and manufacturing data. 

Check out news coverage of the 2025 Data Science Challenge!

What it takes

• Undergraduates interested in data science or related disciplines
• Graduate students experienced in research or applying skills to a research environment
• Students actively pursuing a degree in mathematics, computer science, engineering, science, or other relevant fields
• Students with computational experience

What to expect

Don’t miss this unique opportunity! The experience will be unlike any other. Read about DSC mentors and past Challenges.

This intensive, two-week full-time program provides challenging exercises and assignments, tours, and seminars. For two weeks, you’ll work on an important data science problem while learning from experts, networking with peers, and developing skills for future internships. You’ll also get a taste of day-to-day life at LLNL, where we have a passion for national service. The Challenge will take place in Livermore with a stipend provided.

Team members will:

• Work with scientists, engineers, and technical staff to apply skills to the challenge problem in computational science, numerical methods, mathematics, and science
• Gather and analyze data in support of scientific research
• Participate in research and challenge problem evaluation discussions
• Present results to scientists, engineers, and technical staff during final student briefing
• Actively participate in team meetings and large group meetings

Team leads will:

• Guide the research direction of 4–5  students 
• Attend team lead meetings 
• Lead presentation of results to scientists, engineers, and technical staff during final student briefing 
• Create and schedule meetings for your team 
• Provide advanced technical support to scientists, engineers, and technical staff in materials research and development 

What to include in your submission

A complete submission consists of the following:

A. Resume: make sure to include any relevant coursework. 

B. Letter of recommendation by a faculty member: this person should be able to speak to the quality of your work and potentially as your role as a team member or team leader if you are interested in that position.

• Optional for undergraduates (unless applying to be a team lead then it is required) and is required for graduates. 
• Graduate students: The recommending faculty member should be your graduate advisor. 
• The letter of recommendation  must be sent on your behalf to DataScienceChallenge [at] llnl.gov (DataScienceChallenge[at]llnl[dot]gov) (subject line: "Recommendation Letter <Your Last Name>") 
• Letters of recommendation are due by March 15. 

Please note:

• This program is only open to UC Merced and UC Riverside at this time. 
• Students who have graduated within one year of the Summer 2025 program are still eligible to submit their interest.