Maheshwar Murugesan didn't have the same interests as his classmates in high school.

Sure, he was on the club soccer team, he played the drums, he gave nicknames to his friends – regular high school things. But before Mahesh finished senior year, he had already founded a company and a nonprofit, written papers on nanotechnology and celiac disease, published a book about personal finance, and used scanning electron microscopes for research into wastewater sustainability.

Mahesh, now 18, is the CEO of Effluent, a startup implementing AI into wastewater management. With his cofounder and best friend, David Wang, they are rethinking how water treatment facilities run. Effluent is building tools to help plants stay better aligned with EPA and Clean Water Act requirements, catch biological problems early, and prevent facility mechanical failures before they happen.

Their goal is to create a future with fully autonomous municipal water infrastructure.

Begging for a corner of a lab

Before any of that, Mahesh was a Tamil kid from Cary, North Carolina. He loved sports and music, and he was a good student in school. He was also an avid fan of e-bikes, which is what inspired his first company: Caryeleccycles. Starting his sophomore year of high school, he began retrofitting ordinary bicycles with motors and batteries.

Mahesh’s father is a researcher at Duke University, twenty minutes down the road in North Carolina’s Research Triangle. The apple didn't fall far from the tree. Aside from Mahesh's small business and high school extracurriculars, Mahesh began branching off from the projects he was assigned in school to work on research of his own.

"I was waking up before school to go to the lab. I was going to the lab after school," he said. His method for picking what to study was simple: if a friend was dealing with something, Mahesh wanted to fix it.

When a friend had celiac disease, the two of them came together to write a paper on detecting celiac from the gut microbiome using machine learning models. "I'm down to do anything where I can learn a lot," he said. "If something matters to someone else, I typically like to help."

One of those favors turned into a company. It started when Mahesh set out to help David, whose grandfather passed away from drinking water contaminated with lead, a failure of wastewater management in China.

The two met freshman year at club soccer tryouts. David showed up in a Beijing FC jersey, and since Mahesh didn't know his name yet, he spent the day yelling "Beijing, pass the ball!" They didn't cross paths again until months later, when David walked into Mahesh's biology class. "Yo, Beijing!" David took the seat next to him, and then in every class after that. They were fast friends.

"He's honestly brilliant. He completes me as a technical founder," Mahesh said. "I'm more of a proactive person, I'm more of an optimist. He's more of a pessimist, and we come together to find solutions. He's my soulmate, my best friend and my cofounder."

It was David who first proposed they team up to start researching wastewater, with a focus on cheap, effective treatment for regions with less developed water infrastructure.

"We went across North Carolina and begged people for a corner of a lab," Mahesh said. Working in whatever borrowed space they could find, the two began developing a biochar-silver nanoparticle composite filter.

The filter was fabricated almost entirely from waste. First, they created biochar by heating discarded biomass (they used scrap wood chips) with very little oxygen, a process called pyrolysis, which leaves a lightweight carbon riddled with microscopic pores. That porous structure gives it a huge internal surface area, so it behaves like a sponge for contaminants: heavy metals such as lead and chromium, along with toxic dyes, cling to it and get pulled out of the water.

To deal with bacteria, they use orange peel extract as a natural reducing agent to turn dissolved silver into silver nanoparticles, then bond those particles onto the biochar. Silver is naturally antimicrobial, so the coated biochar also kills microbes like E. coli and Staphylococcus aureus on contact.

Packed into a housing made from reused plastic, the result is one cheap, reusable material that strips out metals, dyes, and bacteria from water.

But they needed to confirm the silver nanoparticles had formed and attached properly. They approached North Carolina State University's Analytical Instrumentation Facility to ask for time on its scanning electron microscopes. Once the AIF saw how much the two had done with so little, it gave them a grant, making Mahesh and David the first high schoolers to receive one from the department.

Their research won North Carolina's regional science and engineering fair, then the NC One Water Award, then the coveted Stockholm Junior Water Prize.

Striking while the iron is hot

Mahesh and David became young leaders in a field almost no teenager thinks about. Mahesh, for his part, was an acclaimed researcher and business owner. But they were still in high school, and college was the expected next step.

Mahesh went to attend his first semester at the University of Michigan's Ross School of Business. He loved research, but he felt a stronger pull toward entrepreneurship. The instinct showed up everywhere. He wrote The Wealthy Teen in high school, a plain-spoken guide to personal finance for people his own age. He hosted a podcast, MM Interviews, where he brought on founders and business owners to tell their stories.

He quickly involved himself with the Michigan Venture Club after arriving on campus. He also started a startup community on campus called The Syndicate to connect students with venture capitalists.

David, meanwhile, was preparing for Duke's Pratt School of Engineering, to pair his scientific interests with technical proficiency. It looked like the two would put their ambitions on hold for a degree.

But when Michigan cut down the nets for the national championship last spring, Mahesh wasn't in Ann Arbor to see it. When Duke's season ended on a 35-foot buzzer-beater by UConn's Braylon Mullins in the Elite Eight, David was preoccupied with something else.

David, who was getting ready for Duke, couldn't stop thinking about the possibility of starting a company. When he could have been enjoying the summer months before school started, he kept building.

Then, Mahesh took leave of absence from Michigan after his first semester. As Mahesh describes it, the next five years are the window that matters – when all the change will happen. School will still be there afterward.

So instead of staying in school, they moved to San Francisco to create Effluent.

Rethinking wastewater, one plant at a time

Effluent is building the intelligence layer for water infrastructure. The company is giving treatment facilities modern tools to analyze their data, better resources to treat water more efficiently, and helping plants prepare for a new era of municipal improvement that will require better technical infrastructure.

Mahesh doesn’t mince his words when it comes to the potential impact of Effluent’s technology: "Wastewater is one of America's top three key infrastructures. You don't understand – the state of this country's wastewater was screwed," he said.

Effluent takes aim at five problems in how wastewater plants run.

First: A typical wastewater facility runs on five separate legacy systems that don't talk to each other, leaving managers to manually stitch the data together. Effluent's first job is to pull the data from all five sources together to simplify how operators can make decisions.

Second: Every wastewater treatment facility is required to file discharge monitoring reports with the EPA. These reports are created by pulling data out of systems called LIMS and WIMS, reconciling that data against paper records, and then submitting the DMR through the federal reporting system. Done by hand, it can take 40 hours or more. Effluent can generate and file those reports in minutes.

Third: The Clean Water Act, passed in 1972, created a wave of new jobs and trained an entire generation of facility operators. That generation is retiring: the EPA estimated in 2020 that roughly a third of the country's drinking water and wastewater workforce would become eligible to retire within the decade, with a median worker age of 48 and only 10% under the age of 24. Their intuition is walking out the door with them. Effluent tries to preserve some of it by recommending decisions based on how a facility's operators have handled similar situations before.

Fourth: Wastewater facilities function with hundreds of interconnected sensors that collect data on water quality, environmental conditions, biological activity, and more. If a plant can't trust its sensors, its operations are crippled. And the biggest issue Mahesh and David have found is that, for the most part, facilities are helpless to prevent sensor failures; they can only respond to failures after they've happened.

One catastrophic outcome of a sensor failure is killing off bacteria cultures. Treatment plants rely on living bacterial cultures to break down waste, and keeping those cultures alive means holding air, temperature, and humidity within narrow bands. One bad sensor reading in the chamber can wipe out an entire culture. Because the biology behaves like a black box, Effluent uses a graph neural network to flag anomalies before they cascade into failures.

Fifth: Improved sensor monitoring offers downstream benefits for facility management. Sensors slowly drift out of calibration and eventually fail, but workers generally only discover a replacement is needed once a sensor is already dead. Since replacement parts for specialized treatment sensors can take weeks to arrive, the lag between a system failure and a receiving replacement can be a major impediment. Effluent watches calibration trends, predicts failures before they happen, and routes replacement orders through the facility's existing, compliant ordering channels so the new part shows up before the old one breaks.

The endgame Mahesh describes is fully autonomous, AI-native water infrastructure: wastewater plants and drinking water utilities that can largely run themselves. The Effluent team is already moving beyond software into hardware, building gas-sensor drones designed to detect emissions from a facility's chemical readings.

The right time to bet on something boring

Effluent arrived in Silicon Valley at the right moment.

Artificial intelligence pulled in about 61% of global venture capital in 2025, yet investor interest has been shifting away from chatbots and toward the physical, regulated, unglamorous corners of the economy that software mostly skipped: manufacturing, energy, supply chains, heavy industry.

Y Combinator, the accelerator that helped define the latest era of startups, now openly a sks founders to take on industrial projects, aging enterprise software, and the machinery that runs factories and utilities. One investor's shortlist for what is fundable right now put "cross-industrial AI" first, climate tech second, and defense tech third.Even in a cooling climate market, startups that pair environmental problems with AI have drawn record funding.

Mahesh rejects that he timed any of it. He's betting the field stays boring. "If you're riding a trend, it's too late. Wastewater is never going to be a trend," he said.

He has a point. Mahesh didn't choose wastewater because it was fundable. He chose it years before anyone was writing checks for it because he wanted to help a friend in need by applying the science he knew better than almost anyone his age. It wasn’t luck, it was persistent dedication to a goal.

For now, Mahesh, David, and the Effluent team are heads-down in San Francisco turning borrowed-lab experiments into a company. But when Mahesh explains why the pieces fell the way they did, he doesn't talk about market timing or the years he spent researching and business-building. He talks about faith.

"God blessed us with this perfect opportunity. Everything in my life perfectly fit into this moment,” he said. “Wastewater research, my best friend going through this difficult time, me having more knowledge about this subject than anyone my age and being in touch with AI. It's been amazing.”