From Sleeping on an Air Mattress to a $275M Exit: Inside Dustin DeVan’s Second Act with Ediphi

Dustin DeVan slept on an ex-girlfriend’s floor on an air mattress at 29. He sold his car for $5,000 to keep the lights on. He managed a dormitory full of 60 foreign exchange students just to get free rent, and signed an affidavit as a murder witness because the dorm cameras caught what happened. He was so broke during the build of his first company that one of his seed investors quietly arranged the dorm gig because he had nowhere else to live.

Three years later, Autodesk wrote a check for $275 million for the company he built off the back of all that, BuildingConnected. Announced on December 20, 2018, the deal made BuildingConnected, with its network of more than 700,000 construction professionals, the largest digital network of construction professionals in the world. It closed on January 23, 2019.

What makes Dustin’s story matter for anyone building, buying, or backing construction technology is not the exit number. It is the four lessons hiding inside it. Why he should have failed and did not. Why the acquisition that made him rich also broke his heart. Why he thinks pre-construction is greenfield. And why no one should ever try this alone.

The solo founder trap

Why doing it alone almost killed the business

Dustin spent the first nine months of BuildingConnected as a solo founder, paying a team of Latvian developers $250,000 to build the product based on ten wireframes. He had $5,000 in his bank account when he signed the contract. He quit his construction job the next day. He raised $75,000 from two friends, another $30,000 from a third, $12,000 from a fourth, and then walked from coffee shop to coffee shop between Palo Alto and San Francisco pitching strangers with laptops who looked like they might have money. That tactic raised him another $125,000.

The code that came back was, in his words, dog shit. He told the developers he could not make the final payment, took the prototype, and used it to recruit Jesse Peterson, a friend from college working at another startup. Jesse threw every line of code away and started again from scratch.

“Doing a startup alone is far too isolating and emotionally taxing to not have someone that you can talk to every day about the problems you’re going through,” Dustin says. He is blunt that starting solo was the biggest mistake he ever made, because it cost him real money and a year of his life chasing a build that was never going to work. The mental health argument matters here too. Silicon Valley has watched 15 years of founders self medicate with drugs and alcohol because they had no one to share the load with. A co-founder is not just product leverage, it is emotional regulation.

The network thesis that built a unicorn

Why most ConTech software is architected wrong

When Dustin sat down in 2011 to write a thesis for what could change commercial construction, he landed on two architectural truths that most software at the time ignored. First, most applications were still desktop tools that had been clumsily ported to the cloud. Second, and more importantly, none of them were built for the reality that construction is highly collaborative yet highly fragmented. Every project is a different assembly of companies coming together for one shared outcome, but the software treated each company like an island.

His answer was a free invitation tool. Other platforms were charging for it. He gave it away because the real prize was the network underneath. Every subcontractor who responded to a bid had to create a profile. That profile became a node. Stack enough nodes together and you have a defensible network that nobody else can replicate without years of work. It is part of the reason Autodesk had to buy them: replicating that network internally would have taken years.

This same logic, network as moat, is exactly what Ediphi is now repeating in pre-construction. Norwest Venture Partners led Ediphi’s $12 million Series A in 2024, after looking at more than 50 construction software companies. The thesis they backed is that the entire industry has been pricing projects backwards for decades. You design first, then price. Dustin’s argument is that you should price the idea first and let cost inform design, because pricing iterations are an order of magnitude faster than design iterations.

If he is right, the productivity gain is not 10 or 20 percent. It is compressing pre-construction from years to months. For the Digital Pragmatist running tight margins, that is the difference between landing a contract and losing it to someone faster. For the AEC investor, it is a thesis worth understanding before the next round.

What goes wrong after the term sheet

The acquisition that broke its own product

Founders dreaming of the big exit rarely hear the second half of the story. Dustin’s is worth sitting with. The $275 million headline number is real, but the deal had clawback clauses, payment schedules, and a 30 day SEC approval delay. The closing party happened before any money hit his bank account. A friend had to pay the bar tab.

The harder part came after. Autodesk had just bought PlanGrid for $875 million a month earlier. When the two companies merged inside Autodesk Construction, leadership decided to reorganise the sales team so every rep sold every product. Dustin stood up in a leadership meeting and called it one of the dumbest ideas he had ever heard, because BuildingConnected had a high frequency, low ACV sales motion aimed at pre-construction, PlanGrid was selling to the field, and BIM 360 was selling to the enterprise. Three different ICPs, three different motions, three different deal sizes. One unified quota.

He was right. BuildingConnected’s sellability dropped dramatically because reps chased the bigger ticket items to hit quota. The product Dustin built became, in his words, a “second tier child”. He stayed for the three year non-compete, watched Autodesk fail to invest in pre-construction, then left and built Ediphi to attack the exact problem he had told them was greenfield. Many of his old team came with him.

The pre-construction land grab

Why Ediphi is Dustin’s bigger bet

Dustin is on record saying he is more bullish about Ediphi than he ever was about BuildingConnected. That is a wild claim from someone whose first company got bought for $275 million. The reason is market structure. Roughly three-quarters of construction estimators are still using spreadsheets, while about a quarter use aging on-premise software. The legacy incumbents have not been innovating. The cloud has not really arrived in pre-con. And the discipline itself sits at the most consequential moment of the project, where most of the cost gets locked in.

The Tech Operator persona will recognise the playbook. Cloud native, network effects, real time collaboration, historical data as a moat, ICP defined narrowly enough to do account based marketing. Early adopters of modern pre-construction tools have reported up to 25% more revenue per estimator. Suffolk Construction’s National Director of Preconstruction Scott Menard ran a study and reported a 1,900 percent ROI, with revenue per estimator increasing by $20 million. Those are not vendor numbers from a marketing deck. They came from Ediphi’s first customer.

For the construction exec staring at 2 to 3 percent margins and a labour crisis, the case is even simpler. If pre-con compresses from years to months, your bid win rate goes up, your re-design loop shrinks, and your best estimators stop drowning in Excel.

How Ediphi compares to legacy estimating tools

The shift from event to process

The pre-construction software market is split between spreadsheets, legacy desktop tools, and the new wave of cloud native platforms. Here is how the three categories stack up on the dimensions that actually matter to a head of pre-con.

The founder mindset that makes the second time easier

Why Dustin is doing it sober and with his old team

One detail from the conversation does not get enough airtime in founder content. Dustin does not drink anymore. He cut alcohol and the rest of it because he wants to give his investors, his employees and his customers the best version of himself. He suffered through depression at BuildingConnected and used alcohol and drugs on weekends to drown it out. He is open about that, and open about the fact that those choices slowed him down.

The second time founder advantage is real if you treated your first company’s people well. Most of Ediphi’s early hires came from his BuildingConnected team. His first hire was a customer support lead, Javier, who started as a college student at BuildingConnected and joined Ediphi years later because he trusted Dustin. That kind of recruiting compound interest is not on any pitch deck, but it is the difference between hiring fast and hiring right.

His advice for anyone thinking about starting a ConTech company is unsentimental. The likelihood you fail is high. Do it for the learning, not the money or the glory. Find a co-founder. Be specific in your positioning, not generic. Be cautious about which VCs you take money from, because the AI hype cycle has put a lot of bad capital into circulation. And remember that AI is at best a feature inside a product, not a product itself.

Frequently asked questions

Related Articles
How BuildingConnected Was Sold for $275m + NEW Super Series
Inside Ediphi: 10 Game-Changing Lessons for Construction Tech