MACRO — Machine Learning-Augmented CRISPR Reprogramming Optimization — sits at the junction of two fields that have only recently begun to combine. Metabolic engineering aims to redirect a cell's internal chemistry toward a desired product; CRISPR provides the editing precision to attempt it; and machine learning offers a way to search the vast space of possible edits without testing each one at the bench. In practice this means the model must predict a system-level outcome — total lipid productivity — from a genetic change, which is exactly the kind of nonlinear, multi-objective problem where computational prediction earns its keep.
She was unusually explicit that the hard part was the point. "I entered because I wanted to take an idea stuck in my head and turn it into something more serious," she wrote in her finalist profile. "I kept thinking about microalgae biofuel because it sounds so promising: algae can grow quickly, absorb carbon dioxide, and produce oils that can be used as fuel. But then I learned that making a biofuel system actually work is much harder than just saying 'use algae.' That made the project more interesting to me."
The problem she selected is the one the field has not closed. "I wanted to focus on the hard part which is how to increase oil production without hurting growth," she wrote. Her reason for choosing this competition in particular was about communication: it "seemed like the right place to test whether I could explain that idea clearly and make it matter to people who may not already know about microalgae or CRISPR." Judging in the 3M Young Scientist Challenge weighs communication effectiveness alongside creativity and scientific knowledge, so the test was a real one.
Her favourite invention of the last century is CRISPR-Cas9, and her account of why is a small essay on where good ideas come from. "Before learning about CRISPR, I thought of DNA mostly as information. CRISPR made me realize that DNA could also be edited in a much more targeted and meaningful way," she wrote. "The part I find most interesting is that CRISPR came from bacteria. It was not invented out of nowhere. Scientists noticed a system bacteria already used to defend themselves and turned it into a tool for research, medicine, agriculture, and biotechnology. I like that because it shows how many useful ideas might already exist in nature, but we have to notice them first."
Her assigned 3M mentor is Aditya Banerji, a Research Specialist in the company's Corporate Research Process Laboratory who develops and commercialises products and is a co-inventor on sixteen 3M invention submissions, four of which have resulted in pending patent applications. For a project whose central risk is the distance between a predictive model and a working process, a process-research specialist with a commercialisation record is a precise fit.
Her ambition is dual by design. "In 15 years, I hope to be a physician scientist," she wrote. "I want to work with patients as a doctor, but I also want to keep asking research questions because I like the part of science where the answer is not obvious yet. I am interested in biology because it is never as simple as I first expected. There is always another layer." She added a wish that has nothing to do with credentials: "I also hope I am still the kind of person who gets excited by a strange question and follows it further than I planned." Her chosen quotation is her own: "A good invention starts when something promising still has a problem nobody has solved well enough." On 12–13 October 2026 she presents MACRO at the 3M Innovation Center in St. Paul, Minnesota.
“I wanted to focus on the hard part which is how to increase oil production without hurting growth.”— Millie Pradawong, 2026 3M Young Scientist Challenge profile
“A good invention starts when something promising still has a problem nobody has solved well enough.”— Millie Pradawong
| Person | Country | Milestone | Age / Stat |
|---|---|---|---|
| Millie Pradawong | 🇺🇸 USA | MACRO — machine learning for CRISPR optimisation of algal biofuel | Age 14 |
| Sharvi Mahajan | 🇺🇸 USA | NeuroDrive Alert — EEG machine learning predicting microsleep | Age 14 |
| Matteo Paz | 🇺🇸 USA | Machine learning on infrared survey data; 1.5m objects found | Age 18 |
| Anirudh Rao | 🇺🇸 USA | Graphene-oxide moisture-powered nano-generator | Age 14 |
Most students who take on algal biofuel take on the version that sounds exciting. She identified the specific reason the field has stalled — the yield-versus-growth trade-off — and built a tool aimed at that, which is the difference between choosing a topic and choosing a problem.
Her reading of CRISPR's origin is the more revealing answer. She fixed on the fact that it was noticed in bacteria rather than designed from scratch, and drew the general lesson: that useful mechanisms may already exist and the scarce skill is attention. That is a research philosophy, arrived at by a fourteen-year-old.
Discover More Child Prodigies
50+ stories of contemporary young geniuses changing chess, music, science, and art.
Explore All Prodigies →
.jpg)
