He runs a New York software company called Assembly. Until nine months ago it was called Copilot. It sells the boring layer between a service firm and its clients - portals, invoices, onboarding forms - and roughly a million people log into it without ever wondering who built it.
Assembly is what you get when a physics-and-business dual major at Western University's Ivey school spends a few years as a product manager at The Climate Corporation, tries two startups that do not stick, gets into Y Combinator in the winter of 2018, and then in the spring of 2020 - which is not, on paper, a promising quarter to start a company - decides that the thing worth building is the client portal.
The client portal is the screen your accountant emails you a link to. Or your lawyer. Or your web agency. It is where you upload the tax document, sign the contract, pay the invoice, and check whether anyone has looked at the tax document since you uploaded it. Historically the client portal has been the least loved surface in professional services software - the part firms tolerated, not the part they marketed. Marlon Misra and his cofounder Neil Raina built a company on the theory that this was, in fact, the entire product.
They called it Copilot. Then, in September of 2025, they called it Assembly. The pitch did not change. The market moved. When the company launched, calling yourself a copilot signaled a small useful tool that sat beside a professional. By 2025, "copilot" had been eaten by every AI company on earth, and the name was doing less work than the product deserved. Assembly is a better word for what the platform actually does, which is bolt together the ten or twelve pieces of software a small services firm needs to run - CRM, billing, files, messaging, forms, contracts - and hand them back as one branded surface.
Misra's public description of the product, delivered on his personal blog and in a dozen interviews, is that Assembly is trying to be "Shopify for service businesses." Shopify is a helpful reference class here. It didn't invent the storefront, and Assembly is not inventing the portal. What Shopify did was decide that the merchant, and not the buyer, was the customer, and then obsessively make life easier for that merchant. Assembly has made the same call: the firm is the customer, the firm's clients are the beneficiaries, and the firm's willingness to pay a monthly SaaS bill for something that used to be a Dropbox folder and a Stripe link is the business.
By the end of 2025 that thesis had produced roughly a thousand paying firms and something in the neighborhood of a million of their end clients, which is a strange kind of scale. Assembly is not a consumer product, but a million consumers have logged into one, without ever hearing the company's name. Capital One, of all things, uses it for a team inside the bank. That is either flattering or worrying, depending on how you feel about enterprise procurement.
Misra's resume reads like a founder who ran the experiments in order. Product manager at a large ag-data company. Then a first startup, Piccolo, where he was CEO from 2017 through 2020. Then a short second startup, Lightout, which measured digital out-of-home advertising - a real business, but a market with a ceiling. Then Copilot, now Assembly, which is neither ag-tech nor ad-tech but is the one that stuck.
The interesting part is what carries across. His first job at The Climate Corporation was product; he left with a reputation, per former colleagues, for "natural product instincts" and "strong technical skills." He picked up a Machine Learning nanodegree from Udacity and, separately and slightly oddly, a Self-Driving Car engineering nanodegree from the same program. Then he built a company whose actual product surface is very deliberately not flashy - forms, files, invoices, chat - but whose underlying bet is the same one he was making in ML class: that software will keep eating the pieces of work that used to require a person on the phone.
"Startups, technology, science, economics and whatever other topic I'm fascinated by in the moment." His personal blog and Substack are both self-titled and both still active while he runs the company.
On September 30, 2025, the company Misra had spent five years building announced it was called Assembly. The press release described it as "the first AI-powered platform purpose-built for professional service firms" and introduced an AI assistant designed to sit inside the client portal - drafting messages, chasing missing documents, summarizing threads. It was the largest release in the company's history.
Renaming a company at Series A is a decision that costs money, breaks links, confuses customers, and forces a founder to explain the same thing to the same reporter twice. Founders do it anyway when they believe the old name is limiting what the product can become. Assembly is a bigger word than Copilot. It implies something you build with, not something that follows you around. Given that the pitch is now Shopify for service businesses, "assembly" does a lot more work.
The domain, assembly.com, is a real one, which is another small tell. A company willing to spend on a five-letter dot-com is signaling a longer time horizon than a company holding onto a Series A budget for growth ads.
Sold to agencies, accountants, consultants, law firms, marketing shops and, in at least one case, a team inside Capital One.
Founders in Misra's cohort tend to specialize. The best-known Y Combinator W18 companies were sharply defined: infrastructure, fintech, dev tools. Misra kept a broader posture. His personal domain redirects to a Substack that covers startups, technology, science, and economics; the "About" page still lists his contact address as marlon at copilot dot com, a small artifact from before the rebrand that he has not gotten around to fixing.
The stated ambition is to lower the operational floor of running a professional services firm to something a solo operator can clear on a Sunday afternoon. If Shopify made "start a store" a checklist rather than a project, Assembly wants to do the same for "start a consultancy" - which today still involves a Stripe account, a Docusign account, a Google Workspace, a project tool, a CRM, and a Notion page that nobody clicks.
There is a secondary aspiration, visible mostly in Misra's writing rather than his product announcements, which is that if you make it easier to run a small services firm, more people will start one. This is the "unlock more entrepreneurship" line that shows up in his LinkedIn headline. It reads like marketing until you notice that Assembly's growth loop actually depends on it: the more firms exist, the more portals ship, the more end clients log in, the more brand equity accretes to the product they never quite noticed they were using.
Copilot officially becomes Assembly. Company unveils its first AI assistant purpose-built for professional service firms, in what it calls its largest release since founding.
Appears on the Software Leaders Uncensored podcast to discuss Assembly, lessons from Y Combinator, and building "the client portal of the future."
He completed both the ML and Self-Driving Car engineering programs at Udacity - unusual credentials for a SaaS CEO, and a hint at what he thought the next decade would look like when he picked them up.
Physics and Business at Ivey. The dual-major move that generally signals a student who could not decide whether to be an engineer or a founder, and so decided to be both.
He owns @asm on Twitter as well as @marlonmisra. The three-letter handle is the ticker of the new company, six months before anyone else would have thought to grab it.
The cofounder and CEO of Assembly (formerly Copilot), a New York SaaS company that builds client-portal software for professional service firms.
A Y Combinator-backed platform - often pitched as Shopify for service businesses - that helps agencies, accountants, lawyers and consultancies onboard, bill and serve their clients through a branded portal.
Yes. The company launched as Copilot in 2020 and rebranded to Assembly in September 2025 alongside its first AI assistant release.
Western University's Ivey Business School, where he studied physics and business. He also completed the Machine Learning and Self-Driving Car nanodegrees at Udacity.
About $13.16M in total, including a $10M Series A closed in 2023.