Tushar Garg Learned Real Estate by Selling One House. Then He Rebuilt the Move.

The first useful thing Tushar Garg did in real estate was admit he knew almost nothing about it. He had studied computer science in India and Minnesota, researched machine learning before it became dinner-table vocabulary, and spent roughly six years at Microsoft, first working on Bing’s core ranking systems and later as a chief of staff. None of this naturally equips a person to stand in front of a Seattle open house and ask strangers whether they might buy a home through him. Garg did it anyway.

He and Stephen Lane had met at Northwestern’s Kellogg School of Management. Their original class project aimed to encourage energy upgrades in houses. The arithmetic was discourteous: the investment often took too long to pay back, and the market did a poor job valuing the improvements. The pair widened the question. Perhaps the problem was not insulation or a clever thermostat. Perhaps the whole home transaction was badly lit.

So they acquired real-estate licenses, partly to reach the information available to agents and mostly to watch the process from inside. Garg’s former Microsoft boss trusted the novice with a purchase. Other prospects required less elegant courtship. Garg waited near open houses and made his pitch. The technologist became a realtor, which must be among the more expensive ways to conduct user research and among the more revealing.

The fifth-highest offer

One family had spent two years losing bids. When another promising house appeared, Garg called the listing agent and asked a plain question: what did the sellers really care about? Nine offers arrived. His client’s price ranked fifth. It was accepted.

Garg celebrated, then looked at the structure beneath the surprise. A seller is not choosing a number alone. A financed offer carries uncertainty: the appraisal may disappoint, underwriting may wobble, the buyer’s own sale may stall. Cash promises fewer plot twists. If ordinary buyers could make an offer with cash-like certainty, Garg reasoned, the contest might become fairer.

“I never thought I would be a realtor selling houses. However, I knew that acting as a realtor would be a great way to learn about the process.”Tushar Garg

Flyhomes grew from that observation, although its first public idea was more playful: buyers could earn one airline mile for every dollar spent on a home. The reward made a fine headline. The transaction itself remained the serious business. Garg and Lane were still students, taking cheap flights to the Northwest, sometimes flying to Portland and driving north, sleeping on couches and missing classes when a prospective client appeared. In 2015 they helped their first buyer close. Referrals followed.

An engineer in an industry of handoffs

Garg’s route to Seattle had already rewarded detours. Born and raised in India, he studied computer science at IIT Guwahati. An interest in machine learning led him to seek a research internship, then to a professor in Italy. That professor suggested he work with her twin sister at the University of Minnesota. He came to the United States, earned a master’s degree, and eventually joined Microsoft.

The continuity is clearer than the résumé first suggests. Search ranking asks how a system sorts confusing evidence into a useful result. Homebuying asks how a swarm of agents, lenders, inspectors, contracts and deadlines might produce a closing. Both reward attention to dependencies. Housing merely adds moving boxes and the occasional tearful telephone call.

Flyhomes began to coordinate more of those dependencies. It developed cash-offer products, added mortgage services, and created a trade-up program that let owners buy and move before selling the house they already had. The company could front capital or provide guarantees that made an offer less contingent. In 2019, it raised $21 million in equity and $120 million in debt. In 2021, a $150 million Series C followed.

Money made expansion possible; it did not repeal weather. When the housing market abruptly paused in 2020, Flyhomes was holding homes it had bought for clients. The company managed to close those deals. Garg’s public account of the period emphasized resilience and the difficulty of recognizing opportunity while one is still living through a crisis. Then the longer cycle turned. Mortgage rates climbed, transactions slowed, and the proptech boom lost its easy optimism. Flyhomes cut staff in several rounds and moved to a smaller Seattle office.

A company designed to remove uncertainty had encountered an industry-sized dose of it. The response would eventually be a narrower company.

The product that left, and the problem that stayed

For a moment, Flyhomes appeared ready to begin again at the top of the funnel. In 2024 it acquired technology from ZeroDown and launched an AI home-search portal. Garg framed information as the buyer’s first large point of friction. The tool let people ask detailed questions about homes and neighborhoods in conversational language, an engineer’s answer to the listing page.

One year later, Real Brokerage acquired the portal and its related technology assets, along with members of the engineering team. Flyhomes retained and sharpened the financial product. Existing investors then supplied a $15 million Series D to expand wholesale distribution of Buy Before You Sell across the country.

The turn looks less like abandoning technology than deciding where Flyhomes had earned the right to be useful. Instead of employing its own agents and loan officers across the transaction, the company now works through the professionals already advising the customer. Its product addresses the timing knot: a homeowner’s down payment may be trapped as equity in the current home, while the next seller prefers an offer that does not depend on that home selling first.

This is an unglamorous form of strategic maturity. Young companies like verbs such as own, transform and reinvent. Older companies learn the charm of choosing. Flyhomes once tried to provide the search, agent, mortgage, title, escrow and repair experience under one roof. The wholesale model accepts that customers already have trusted professionals. The company can supply a missing financial bridge without asking everyone else to leave the road.

“It’s good to have a plan ... and then you also have to be spontaneous because great opportunities show up.”Tushar Garg

The plan, lightly held

Garg has described his career as spontaneous. The word can sound breezy until one notices how much work his spontaneity demands. A research interest led him to ask around campus until an internship appeared. A classroom project failed its first economic test, so he and Lane inspected the larger market. Real estate was unfamiliar, so they licensed themselves and did the job. A client kept losing, so Garg called the other side and asked what mattered.

Colleagues have described the company as an attempt to operationalize Garg’s concern for people. The phrase is corporate, but the underlying distinction is useful. A property is an asset; a move is a life event with an asset inconveniently attached. The best Flyhomes anecdotes are not about portals or capital facilities. They are about removing the double move, helping a family compete, or recognizing that a customer’s balance sheet may be sturdy while their calendar is impossible.

His technical background gives the story its machinery. His willingness to enter the transaction gives it judgment. Many founders say they are customer-obsessed from a safe distance. Garg’s early version involved open-house pavement, a licensing exam and the possibility of losing Saturday to strangers. He did not merely collect complaints. He acquired responsibilities.

Flyhomes is now a different company from the airline-miles brokerage imagined at Kellogg, and a different company from the integrated platform that raised its Series C. Yet its current aspiration is recognizably the same: let a buyer move with more confidence. Ten years of additions and subtractions have reduced the grand theory to one stubborn piece of choreography. Sell this home. Buy that one. Keep the family out of temporary housing. Make the dates behave.

That focus also changes the measure of the company. A search portal can count visits. A brokerage can count listings. A timing product is judged at the seam between two enormous decisions, where a delayed sale or uncertain offer can rearrange a family’s month. The useful outcome is almost anticlimactic: the old house closes, the new keys work, and the bridge quietly disappears. Infrastructure is often most convincing when nobody has to admire it.

There is wit in the destination. Garg began by investigating why homeowners would not spend money to save energy. He ended up working on a different household conservation problem: how to save time, certainty and one unnecessary move. The class project did not survive. Its curiosity did.