He could have spent his life making chips think faster. Instead he is teaching city hall to pick up the phone.
// Co-Founder & CEO, Polimorphic - New York, NY
Parth Shah's software is answering a phone call right now in a town you have never visited. A resident wants to know whether their dog license renewed, or why the building permit stalled, or how to pay a water bill that arrived on paper. On the other end is not a tired clerk on hold music. It is Polimorphic - an AI front desk that fields the question in any of 75-plus languages, at any hour, and never once sighs.
Shah is the co-founder and CEO. The company he runs out of 122 West 26th Street in Manhattan does something deeply unglamorous and quietly enormous: it builds the first constituent relationship management system designed from scratch for local government. Think Salesforce, but for the county clerk. The unsexy plumbing of democracy, finally given decent pipes.
The pitch lands because Shah refuses to mock the people he is selling to. "Folks underestimate how critical a lot of local government and state agency work is," he says. "It's not the public servant's fault. They're trying their best. They just haven't had the right systems." That sentence is the whole company. Empathy as a go-to-market strategy.
In July 2025 the bet got bigger. Polimorphic announced an $18.6 million Series A led by General Catalyst, with M13 and Shine returning. Total raised: roughly $28 million. The money goes where the mission does - more salespeople, more support staff, more states. Less waiting in line.
"Why is it easy to track an athlete's stats, but not a politician's?"- The question, asked in an MIT dorm, that started everything
Rewind to MIT. Shah and his co-founder Daniel Smith were first-years who turned their weekends into a salon - informal sessions where they argued about the problems facing the world. Shah kept circling back to government, and to educational policy in particular. One frustration nagged at him: you can pull up an athlete's every statistic in seconds, but try finding out what your representative actually did last week.
So they built the obvious thing. A civic-data platform - call it ESPN for politics - to make political activity as legible as a box score. They sharpened the concept through the MIT Sandbox innovation program, launched a beta senior year, and after graduation joined a summer accelerator and landed funding from Pear VC.
Then they did the harder thing: they pivoted. The team realized the real bottleneck was not transparency at the top but capacity at the bottom - the overwhelmed town offices where democracy actually touches people. Polimorphic became a tool for governments themselves. The scoreboard became the front desk.
Shah came to the problem honestly. He says he learned how government works behind the scenes from his grandfather, a local-government utility director, and his grandmother, who spent a career in state government. The dinner-table education turned out to be a market map.
Figures reported by Polimorphic across customer governments serving 1,000-100,000 constituents.
The numbers read like a magic trick, but the mechanism is mundane. Most of what overwhelms a government office is repetitive: status updates, form intake, the same five questions asked a thousand ways. Shah has a favorite metaphor for it - the "laundry and dishes" of public service. Necessary. Endless. Soul-flattening.
Hand the laundry to AI and something opens up. Voicemails drop by up to 90%. Walk-ins fall by three-quarters. Across customers, Polimorphic has helped collect more than $10 million in online payments and clawed back over 55,000 working hours - about 26 years of human labor returned to the actual job of serving a community.
It is not abstract. In Vernon, a Polimorphic voice line named "Trudy" now brings town hall to residents around the clock. The future of government, it turns out, has a first name.
It's not the public servant's fault. They're trying their best. They just haven't had the right systems.
Folks underestimate how critical a lot of local government and state agency work is.
Why is it easy to track an athlete's stats, but not a politician's?
The name Polimorphic winks at both "politics" and "polymorphic" - software that takes many forms.
Before chatbots for city hall, Shah was tuning deep learning models at NVIDIA.
His turf is small-town America: customers serve anywhere from 1,000 to 100,000 residents.
His grandfather ran a local utility; his grandmother worked in state government. The job was practically inherited.
One deployment, a voice assistant named "Trudy," literally answers town hall's phone overnight.