A residential appraisal contains a peculiar little tragedy. The difficult part is judgment: deciding what a property is, how its condition compares, and what the evidence supports. Yet the appraiser can lose hours to the easy part, measuring rooms, sorting photographs, drawing a floor plan, and entering the same facts into one box after another. Anir Prativadi has built his young company around that mismatch. ValueMate begins with an iPhone and a walk through a house. Its wager is that the phone can remember the walls, organize the notes, and prepare the forms so the professional can spend more time being professional.
Prativadi is the co-founder and CEO of ValueMate, a New York company in Y Combinator's Spring 2025 batch. The product uses phone-based LiDAR to capture a property during a roughly three-minute walkthrough. From that pass, it creates a 3D model and a two-dimensional floor plan, gathers audio and video notes, and drafts much of the appraisal report. Square footage, condition codes, narrative, and the new UAD 3.6 fields can arrive mostly populated. The appraiser reviews the result, makes the judgment calls, and signs the work.
The sequence matters more than any single technical ingredient. LiDAR already exists in a phone. Generative AI already writes fluent paragraphs. Floor-plan software already draws straight lines. ValueMate's work is to make those pieces behave like one continuous thought. Capture a fact once, carry it forward, and stop asking a human to retype it because two pieces of software have never been properly introduced.
A founder story assembled in dorm rooms
The company began with a friendship. Prativadi met Pietro Demicheli at Carnegie Mellon University, where both studied artificial intelligence. They became friends before they became co-founders. The pair had personally encountered real estate transactions slowed by appraisals, then went looking for the blockage. The blockage was not simply a long inspection. It was the sediment of administrative work that settled afterward.
There was another Carnegie Mellon relationship in the wings. As a freshman, Prativadi shared a room with Vivek Isukapalli. He watched Isukapalli build a company and scale it to seven figures, an experience Prativadi later described as making entrepreneurial ambition tangible. Y Combinator had been a dream; now someone on the other side of the room had demonstrated the route. After Prativadi and Demicheli entered the program with ValueMate, Isukapalli joined them as a co-founder.
“Getting into Y Combinator has always been more than just a goal.”Anir Prativadi, on welcoming Vivek Isukapalli to ValueMate
The anecdote reveals something cleaner than the usual mythology of a solitary founder receiving a lightning bolt. Ambition can be social. One friend makes an unfamiliar outcome visible. Another becomes the person with whom a narrow industry problem can be examined for months. The founding team is held together by an institution, a field of study, and the more useful bond of having known one another before the company had a batch name.
Prativadi's path to appraisal technology also contains a stylistic detour. Before ValueMate, he was a profitable music entrepreneur whose ventures reached more than 100,000 people. The transition has a pleasing rhythm: from producing experiences for an audience to orchestrating a production line for property data. Both require disparate parts to arrive together. Appraisal forms are less likely to demand an encore, but they are equally intolerant of missed cues.
The business hiding behind the boring bit
A home appraisal sits inside sales, refinancings, and other transactions. It is consequential and intentionally methodical. That makes the work a tempting target for careless automation and a demanding target for useful software. ValueMate's answer is to put its energy around the licensed professional rather than pretending the professional is a decorative final click. On an August 2026 appraisal podcast, Prativadi emphasized that the appraiser remains in control of judgment calls while the platform reduces time spent on inspection administration and reporting.
This boundary is the product. A model can organize observable characteristics. It can attach a photograph to a room, translate spoken notes into structured text, compute geometry from a scan, and place data in the right fields. The appraiser still decides what those characteristics mean in a valuation. The distinction may feel modest beside grand claims about autonomous work, but regulated markets tend to reward modest distinctions that survive contact with a real file.
Timing sharpened the pitch. Fannie Mae and Freddie Mac have been moving the appraisal industry toward UAD 3.6, a redesigned dataset and reporting framework with more structured property information. A longer, more detailed form is painful when each new field creates another manual action. It is much less alarming when the data travels directly from capture into the report. ValueMate talks about compliance as part of the route, not a chore applied after the route is finished.
Prativadi did something else that makes the story more interesting: ValueMate is paired with VM Appraisals, an appraisal operation built around the same technology. This gives the team a place where product choices meet deadlines, edge cases, and the habits of working appraisers. Software founders adore the phrase “dogfooding.” In property valuation, the meal comes with federal data standards, lender expectations, and a signature carrying professional responsibility. The feedback should be nutritious.
The operator inside the optimist
Prativadi's public summary contains only two words: “Exponential growth.” His posts use the enthusiastic vocabulary of a young founder, with candid declarations of dreams, speed, and belief in his team. Yet ValueMate's actual proposition is almost aggressively concrete. Walk for three minutes. Capture millions of 3D points. Produce a floor plan. Pre-fill the report. Reduce a five-day turnaround to hours. Grand ambition has been given a clipboard and told to show its work.
One small story captures the culture he admires. In San Francisco, Demicheli was due for a driving test when the team realized he needed someone with a California license. Prativadi's Texas license would not do. They called Meaghan, a contact at Brex. She crossed the city and arrived within half an hour. Prativadi presented the rescue as evidence of San Francisco's velocity. It is also evidence that even a LiDAR company can be defeated by the wrong card in a wallet, and rescued by a person willing to answer the phone.
The useful version of speed is not frantic. It is one fact moving cleanly from the room where it was captured to the person who must decide what it means.
By August 2026, Prativadi was carrying the idea into industry conversations. He was listed at Valuation Expo for a mobile-technology demonstration and a panel about valuation technology. He also joined the Appraisal Buzzcast for a discussion framed around AI, UAD 3.6, and the possibility of a 30-minute appraisal. The conversations matter because appraisal software cannot be willed into existence from a startup office. It has to earn the attention of people who know where a report breaks.
There is a lesson here for builders hunting for an AI idea. Begin with the boring after-work, the copying that follows the skilled act. Look for a profession in which the same evidence is repeatedly translated between tools. Connect capture to structure before adding a chat box. And define, with unusual care, which decisions remain human. ValueMate's ambition is large, but its unit of progress is refreshingly domestic: a person walks through a house once.
The house as a dataset, the appraiser as an author
The long-term bet is that a property can become a coherent digital model at the moment of inspection. Its geometry, rooms, observed characteristics, photographs, and notes become parts of the same record rather than loose ingredients assembled later. That model can create a floor plan today, populate a compliant form tomorrow, and support a clearer history of the property over time. Prativadi's aspiration is to build the engine connecting those pieces across the appraisal lifecycle.
Still, a model of a house is not the house. It does not hear the floorboard complain under a careful step or understand why one renovation matters more than another without context. ValueMate's practical stance is that technology should arrive with organized memory, then hand the file to someone qualified to interpret it. The appraiser becomes less of a transcriptionist and more clearly the author of a professional conclusion.
Prativadi is early in both his career and the company's. He was on Carnegie Mellon's Fall 2024 Tepper Dean's List, entered YC the following year, and is now building amid a live standards transition. The evidence so far is a product, an operating firm, early users, conference conversations, and an unusually tight group of friends trying to make a stubborn workflow move. Real estate will always contain walls. Prativadi is betting the paperwork between them does not need to feel quite so solid.