On a Saturday evening in late March 2025, Dushyant Singh Parmar made a panicked phone call to his CTO. Claude AI had published a groundbreaking paper on Model Context Protocol servers back in December 2024, and only a handful of companies worldwide were building context-aware AI agents that could actually execute tasks.
"We can't miss this," Dushyant told Siddharth Bulia, his mentor and Mili's CTO. "This is THE TIME."
One week later, Dushyant won Mili's internal hackathon—a competition designed for builders to create POCs that would shape the company's strategic roadmap. His AI agents' proof of concept triggered a pivot, unlocking 10x growth potential. Two months later, at the Wealth Management EDGE conference in May 2025, Mili beat industry titans like BILL and Addepar for the People's Choice Award.
This wasn't a sudden breakthrough. Years earlier, a successful engineer at one of the world's largest transportation companies had walked away from stability to chase the chaos of startup life.
The Engineer Who Chose Chaos
In 2022, Dushyant had what most engineers dream of: leading teams at Alstom building self-driving metro systems serving 6 million daily riders across Lille and Shenzhen. He supervised teams across two continents, shaping critical infrastructure.
Yet something gnawed at him.
"I wanted to live in the uncertainty around a young startup," Dushyant recalls. "I was inspired by people who built things from scratch—Jordan Mechner creating Prince of Persia, Mark Zuckerberg in a Harvard dorm room, the misfits of Pied Piper from Silicon Valley. I wanted every single decision to directly translate into real-world results."
He missed the raw energy of building from nothing—the all-nighters, the small team racing to change the world, the feature wars where each line of code could make or break you.
So he quit. He pursued an MBA at Bath School of Management, then briefly joined Equivital as Product Manager for wearable health tech. But the startup itch only grew stronger.
Finding the Mission
In March 2024, Dushyant's mentor, Vennela Miryala, was embarking on a new mission with Siddharth Bulia—building Mili to transform wealth advisory. Dushyant knew he had to join her in changing the wealth advisory world for the better.
The founding team came from Meesho—the Indian e-commerce success that had beaten Amazon in tier-2 cities. Their goal? Build the world's best AI note-taking app for a $250 billion market where firms manage trillions in assets. The technical hurdle was massive: create a live AI note-taker that worked invisibly, without recording meetings or using intrusive bots.
"I knew this was my Pied Piper dream calling," Dushyant says.
He joined as a founding AI Product Engineer and Partnerships Lead.
Building v0: The Breakthrough
Every AI note-taking tool worked the same way: a visible bot joined meetings, recorded everything, then generated notes. For wealth advisors handling sensitive client information, this was a compliance nightmare.
"We needed to build something that worked silently in the background, capturing audio in real-time without ever recording or storing conversations," Dushyant explains. "It was our 'Tesla's light bulb' invention—a thousand failed attempts before we got the successful one."
The breakthrough required weeks of late-night debugging. Dushyant architected Mili's entire v0 tech stack from scratch—core functionalities, system integrations, DevOps pipelines, and evaluation systems. He created Sirius—Mili's internal admin panel—from a synthetic dataset generation system to a complete evaluation framework.
Over 5–6 months, he delivered integrations across 10+ enterprise platforms—Salesforce, Microsoft Dynamics, Google Calendar, custodian systems like eMoney and TradePMR. These integrations enabled Mili's wealth advisory clients, who collectively manage over $10 trillion in assets, to seamlessly sync their workflows.
But these weren't just technical pipes. Mili understood meeting context, automatically routing notes to correct CRM objects. Action items are generated in real-time with assignees and due dates pre-filled, syncing in minutes what took advisors weeks.
By December 2024, Mili had secured $2M in funding. But Dushyant's biggest innovation was coming.
The Hackathon That Changed Everything
The insight came from Claude AI's December 2024 paper on Model Context Protocols. By March 2025, Dushyant realized: AI had evolved beyond answering questions. With MCPs, you could build agents that executed entire workflows.
"Note-taking was already saving users 50 hours a month," Dushyant explains. "But there were still so many repetitive touchpoints. That's where the 'Jarvis moment' crystallized."
After that Saturday phone call, Dushyant worked around the clock. He built a proof of concept in one week—using AI to build AI. What would have taken senior architects months, he completed in days.
The POC won Mili's internal hackathon. The company pivoted, establishing AI agents as its primary track.
Dushyant led the development of Mili's first production-ready AI agents suite and built the MCP servers powering them—Salesforce MCP, Wealthbox MCP, Google, and Microsoft Calendar MCPs.
The result? Advisors could command: "Go through all notes from the past 5 years where we discussed crypto with the Johnson Family. Summarize in a one-pager with action items. Send a follow-up email. Schedule a meeting with client snapshot."
Tasks taking days or weeks are completed in under a minute.
Beating the Giants
In May 2025 at the Wealth Management EDGE conference, Mili was the underdog against household names like BILL and Addepar.
In front of hundreds, Dushyant and the team showcased their AI agent performing live deep research on a judge, creating a prospect one-pager, adding him to Salesforce, and drafting onboarding documents—all in 10 seconds.
"Each step would have taken a seasoned advisor days or even a week," Dushyant recalls. "We did it in seconds on stage."
When votes came in, Mili secured over 50% of the audience—more than all competitors combined.
The newcomer with technology that didn't exist anywhere else had beaten the giants.
Building for Impact
The technology matters, but Dushyant's real focus is building products that change how people work.
"Fifteen years ago, none of us would have imagined sharing photos and stories with thousands through a device in our pocket," he says. "Android apps had that impact."
He sees AI as this decade's defining shift—following computers in the '90s, the internet in the 2000s, smartphones in the 2010s. "There's a boom every decade. I didn't want to lose this one. I wanted to list my name among pioneers who left a dent for good."
Innovation at this pace requires sacrifice. "You really need that inner 'why,'" he says. "All of us left high-paying stable jobs. Why? Because we wanted to kick ass. Simple."
Working with Siddharth has been transformative. "My CTO is by far the best leader I've worked with in my 8–9 years," Dushyant shares. "He's incredibly empathetic—he gets what people are feeling and where they're stuck. And when things fail, he doesn't dwell on it. He learns and moves forward."
This was the "Pied Piper moment" Dushyant searched for since leaving Alstom—the band of misfits coming together to change the world.
AI's Leveling Effect
Looking ahead, Dushyant sees AI democratizing entrepreneurship like never before.
"For the first time, we have machines that understand context, learn from context, and execute tasks," he observes. "This is once-in-a-century. That's why you're seeing new inventions every two weeks that used to happen every 2–3 years."
What took civilization 40–50 years now happens in one year. Solo entrepreneurs emerge everywhere, creating value at unprecedented speed.
"We just need to seize the moment," Dushyant says. "All of us can be experts now. AI has leveled the playing field."
Dushyant Singh Parmar's journey shows what happens when deep technical expertise meets bold vision. From building the v0 tech stack to leading the AI agents revolution that won at EDGE, he's proven that the most impactful innovations come from builders who understand both the technology and the problems it solves.
He left stability for uncertainty. Found his mission alongside his mentor. And built the future of AI-powered work, one agent at a time.
ⓒ 2025 TECHTIMES.com All rights reserved. Do not reproduce without permission.
