What this covers

Patrick interviews Max Simkoff, founder and CEO of Doma, moving from the mechanics of a home closing to the mechanics of building a company. Simkoff starts with history: a nineteenth-century Supreme Court ruling on Caveat Emptor (Buyer Beware)) led to the first title insurers. He then lays out what title companies actually do, which goes beyond underwriting to escrow, earnest money, disbursements, and recording. One oddity is that a refinance requires a new title insurance policy because it protects the new lender, so the lender is effectively the customer.

The middle of the conversation covers economics and modeling. Simkoff estimates incumbents spend low-to-mid 80s percent of revenue on operations, giving a mid-to-high 80s combined ratio, and he explains why purchases yield operating leverage that refinances don't. He traces his modeling approach to Evolv, where retention was treated as a survival curve and psychology constructs were paired with econometrics. At Doma, incumbents' completed research files became training data. He reports odd findings, such as bedroom counts predicting defect risk non-linearly, and defends correlation over causation. Computer vision shrinks a 30-minute document task to under three seconds.

The last stretch is about company-building. Simkoff explains why Doma made Lennar a shareholder instead of buying a title company, how reinsurance and balance-sheet capital satisfied Fannie Mae, and why public-market short-termism worries him. He also covers pushing a reluctant board, measuring recruiting NPS, and, prompted by Patrick's reference to Peter Thiel, his own "loose screw" as a founder.

Sharpest takeaway

Simkoff argues that title and escrow work is a labor-heavy, data-poor process that incumbents have never modernized, so machine learning applied to property and closing data can cut cost and speed while outperforming the human 'answer key' approach.

  • Incumbent title companies run manual records research and back-office labor that consumes most gross revenue, with a combined ratio in the mid-to-high 80s.
  • Structured predictive analytics proven at Evolv transfers to title underwriting and closing documents, including unstructured PDFs and computer vision.
  • Regulatory and capital constraints (GSE approval, credit ratings, reinsurance) were the real barriers, solved through partnerships and persistence.

The claims · ranked66 claims · weighted by value

This asset isn't compiled yet

You're seeing its claims, ranked. Compile it to build the argument threads, weight them, and check each claim against your library — the full view.

0.89

Borrowing from epidemiology and biostatistics, Evolv modeled employee retention as a survival curve starting at 100% on day one and tracking the share remaining over time, which captures the continuous path of attrition better than a point-in-time attrition count.

factualhigh valueestablishednovelty 4/4durability 4/4· Max Simkoff

“we should borrow models from epidemiology and biostatistics. And what we ultimately looked at was something that sounds very morbid, but we looked at something called the survival curve.”

0.86

A mid-to-late 1800s US Supreme Court decision established buyer beware, holding that a seller was not necessarily responsible for establishing clear ownership, which led to the first title insurance companies being founded shortly afterward.

causalhigh valueestablishednovelty 3/4durability 4/4· Max Simkoff

“the Supreme Court decided that in this country, it is actually buyer beware. The seller is not necessarily responsible for establishing clear ownership on the property.”

0.81

On a refinance, the borrower must buy a new title insurance policy even though they already paid for one at purchase, because the policy protects the new lender's security interest (a lender's policy) against clouds on title since the last purchase, making the cost largely redundant to the borrower.

causalhigh valueestablishednovelty 3/4durability 3/4· Max Simkoff

“the insurance policy that you're paying for when you do refi is to the benefit of the new lender, as it's called a lender's security policy.”

0.80

For a long time, the primary concern for a startup is not running out of money, a simple point that many founders miss because they fall in love with the idea and forget they are running a Delaware C corporation with shareholders expecting a return.

normativehigh valueestablishednovelty 2/4durability 4/4· Max Simkoff

“your primary concern is don't let the company ran out of money.”

0.75

Incumbent title companies, before Doma, did mini research projects at the County Recorder's and Assessor's offices for virtually every residential transaction to check for outstanding wills or trusts, then 'insured' title by asserting they had looked exhaustively and were near-certain of the new owner.

factualhigh valueestablishednovelty 2/4durability 3/4· Max Simkoff

“they were basically doing these mini research projects to go to the County Recorder's Office and go to the County Assessor's Office”

0.75

Title companies are responsible for closing residential real estate transactions end to end, including escrow setup, earnest money funding, disbursements, paying the realtor commission, collecting signed documents and recording at the county recorder's office, not just underwriting title insurance.

factualhigh valueestablishednovelty 2/4durability 3/4· Max Simkoff

“title companies are responsible for basically closing the transaction, that means everything”

0.75

Different capital instruments serve different goals: preferred equity from venture funds to start the company, balance-sheet capital and reinsurance to win Fannie Mae approval, and a senior secured credit facility to fund a self-funded plan before the public offering.

factualhigh valueestablishednovelty 2/4durability 3/4· Max Simkoff

“There's many different instruments you can use to facilitate many different goals for the company.”

0.74

Purchase transactions are the most profitable in title because they bring roughly 2x the revenue of a refinance, while about 80% of the required steps overlap, giving traditional title businesses substantial operating leverage on purchases.

causalhigh valueestablishednovelty 3/4durability 3/4· Max Simkoff

“a purchase has like 2X the revenue. So you basically get a massive amount of operating leverage”

0.73

Peter Thiel argues that founders' tail traits sit at the extremes of an inverted bell curve, so the quirks or 'screws loose' that look like dysfunction are essential to their success.

factualhigh valuecontestednovelty 3/4durability 3/4· Patrick

“it's actually inverted. It's an upside down bell curve. And so the screws loose are actually essential.”

0.71

By combining industrial-organizational psychologists' constructs such as conscientiousness and punctuality with econometricians' validated statistical models, Evolv could measure traits in an online job application that predicted termination reasons, productivity and promotability before hiring.

causalhigh valueestablishednovelty 2/4durability 3/4· Max Simkoff

“We could predict it all upfront.”

0.69

Based on Simkoff's last look at incumbent P&Ls, low-to-mid 80s percent of gross revenue goes to operating expenses, and with loss ratios of about 3-5% a decent incumbent runs a mid-to-high 80s combined ratio, leaving roughly 10-12 cents of pretax margin, and a bad one in the low 90s.

factualhigh valuecontestednovelty 3/4durability 2/4· Max Simkoff

“low to mid eighties percent of gross revenue goes to OPEX. Then you have to add their loss ratio.”

0.68

Public market investors focused on daily, weekly or quarterly value create pressure that damages long-term returns, so a company should seek investors who remember that stocks like Square and Facebook dipped 20-30% in the months after their IPOs before becoming far more valuable.

causalhigh valuecontestednovelty 2/4durability 3/4· Max Simkoff

“If we pay attention to the value of the company on a daily or weekly or monthly or quarterly basis, we're screwed for longterm returns.”

0.61

The five-year compound annual growth rate of the zip-code-level Housing Price Index from Census data was strongly predictive of certain categories of title defect risk, a signal that no incumbent title examiner would ever manually check.

factualhigh valuespeaker onlynovelty 4/4durability 3/4· Max Simkoff

“We found signal in that dataset on the compound annual growth rate ... the five-year CAGR the zip code level housing price index is strongly predictive of certain categories of title defect risk.”

0.61

In a business predictive model, the team does not need to explain why a variable like bedroom count predicts risk; if the correlation has enough statistical predictive capability, that is sufficient because the company is in the correlation business, not the causation business.

normativehigh valuecontestednovelty 2/4durability 3/4· Max Simkoff

“We're not in the causation business. We're in correlation business. If that thing's got enough statistical predictive capability”

0.60

Being great at predictive modeling begins with precisely defining the outcome to predict, because vague goals like better employment outcomes break down until you specify business value, such as the value of tenure, since good tenure differs from bad tenure.

normativehigh valuespeaker onlynovelty 3/4durability 4/4· Max Simkoff

“Great looks like you have to have a very clear picture of what the outcome is. You have to know what it is that you want to predict.”

0.59

On hundreds of thousands of completed preliminary title reports, the number of bedrooms and bathrooms was linearly predictive of certain categories of title defect risk in some regions up to a point, then inverted, so two-bedroom homes showed more risk than one-bedroom homes in some categories while four-bedroom homes showed less than three-bedroom homes.

factualhigh valuespeaker onlynovelty 4/4durability 3/4· Max Simkoff

“the number of bedrooms and bathrooms in the home was linearly predictive of certain categories of title defect risk to a point, and then they invert.”

0.57

In refinancing, the lender is effectively the customer for title selection because borrowers trust and follow the lender's recommendation, even though the lender does not pay the title cost, so the go-to-market is more B2B than B2C.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Max Simkoff

“the borrower follows the lender's recommendation because they're trying to get a refinance done.”

0.57

Incumbent title companies' completed research files, which they were proud of as an answer key for every transaction, became the training data Doma used to build its underwriting models.

factualhigh valuespeaker onlynovelty 3/4durability 3/4· Max Simkoff

“There was an answer key for every transaction, that title companies were very proud of.”

0.57

Rather than buying a title company, which Simkoff considered highly risky for a startup, Doma found Lennar, a company not trying to sell its title business, and made it a large shareholder so it could supply distribution and data while sharing the long-term vision.

normativehigh valuespeaker onlynovelty 3/4durability 3/4· Max Simkoff

“the reality of M&A is that it's highly risky. Especially a startup buying a large company. It won't work.”

0.57

Doma needed balance sheet capital to satisfy Fannie Mae and Freddie Mac for underwriting title insurance on refinances, so it secured a reinsurer that became a strategic investor and provided a reinsurance agreement, which functioned as a form of financing.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Max

“we found a reinsurer who became a strategic investor in a company. They put cash into the business alongside a reinsurance agreement”

0.57

Accomplished board members often hesitate to challenge a strong CEO, so the CEO must push hard enough to make them lose their composure, which is the signal that the pressure has gone too far.

normativehigh valuespeaker onlynovelty 3/4durability 3/4· Max Simkoff

“I have gotten each one of them to the point where they were screaming. ... that's the only way that I'm going to understand that I'm pushing it too hard.”

0.57

Doma now uses computer vision to classify closing documents by layout and geographic placement of text, such as tables, shading and signature lines, to extract fees, reducing a task that takes a human about 30 minutes to under three seconds.

factualhigh valuespeaker onlynovelty 3/4durability 3/4· Max Simkoff

“when it's a human being doing it, it needs 30 minutes. When it's a model doing it, it needs less than three seconds.”

0.56

Doma measures the NPS of its recruiting process because candidates decide based on first impressions, and recruiting is a sales process where delays are costly, since 'time kills all deals' and strong candidates receive more offers the longer they wait.

normativehigh valuespeaker onlynovelty 2/4durability 4/4· Max Simkoff

“Because recruiting is sales, by the way. Time kills all deals.”

0.54

Most of an incumbent title company's cost of goods sold is back-office labor for records research, fixing title issues and escrow setup, and incumbents spend little on sales and marketing because customer acquisition is not hard in title.

factualhigh valuespeaker onlynovelty 3/4durability 3/4· Max Simkoff

“most of their COGS is back office labor.”

0.54

HR systems recorded termination reason codes, such as involuntary termination for punctuality and attendance, that often mislabeled people who simply stopped showing up as fired, so the raw data required careful interpretation.

factualhigh valuespeaker onlynovelty 3/4durability 3/4· Max Simkoff

“they just stopped showing up for work and we had to term them in the system. And so the code we used was involuntary.”

0.54

Doma hired a COO who was one of the youngest partners at McKinsey, from its digital insurance transformation practice, rather than an industry operator, because the company needed creativity and execution rather than a playing-it-safe, inertia-bound hire from the industry.

normativehigh valuespeaker onlynovelty 3/4durability 3/4· Max Simkoff

“If you just hire people from the industry, you're playing it safe, you're going to be constrained by inertia”

0.54

Doma faced a catch-22 where rating agencies required two years of underwriting history before issuing a credit rating, and Fannie Mae's guidelines required a rating, so Doma got Fannie Mae to accept a ring-fenced pilot, then added reinsurance capital to meet the balance-sheet requirement, and only then obtained a rating.

factualhigh valuespeaker onlynovelty 3/4durability 3/4· Max Simkoff

“So you just... You just need to show us that you've been in business for two years underwriting title insurance.”

0.54

Doma persuaded Fannie Mae that it should be held to the same standard as traditional title insurers, arguing that since incumbents make some mistakes, its model could be allowed to make a somewhat higher number of mistakes if it carried more capital on its balance sheet.

normativehigh valuespeaker onlynovelty 3/4durability 3/4· Max Simkoff

“why would you let them make some number of mistakes and not also let our model make some number of mistakes?”

0.54

At Evolv, applying statistics and econometrics to structured HR data from large employers revealed that hiring outcomes such as tenure and productivity could be predicted from candidates' online application answers, and clients like Xerox, eBay and AT&T cut attrition roughly in half and raised productivity 10-20%.

causalhigh valuespeaker onlynovelty 3/4durability 3/4· Max Simkoff

“Xerox, E-bay, AT&T, they plugged in our solution and over months, or even years of time, they cut their attrition in half. Their productivity went up by like 10 or 20%.”

0.52

Doma integrates directly with lenders' loan origination systems and offers a public API, and its machine accepts inputs such as PDFs and returns them in whatever format the customer wants, so lenders do not need to conform to a new data standard.

factualhigh valuespeaker onlynovelty 2/4durability 3/4· Max Simkoff

“We don't ask them to conform to some new standard of how they pass that information back and forth.”

0.52

Managing stakeholder perception requires retaining long-term confidence while still listening to challenges, and when a team member who doesn't believe in the vision persists, the CEO should tell them they are not a believer and let them leave, disregarding short-term perception.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Max Simkoff

“You're not a believer. And if you're not a believer, this is a bad place for you to be.”

0.52

Large centralized lenders such as Chase, PennyMac and Wells Fargo do not decide on each transaction but allocate a percentage of national volume to a solution provider like Doma if it beats their service metrics, so they can shift volume overnight.

factualhigh valuespeaker onlynovelty 2/4durability 3/4· Max Simkoff

“They basically decide if our solution provider, like Doma, is able to blow away their metrics, then we will send them X percent of our national volume.”

0.52

Accomplished board members who never lose their composure eventually scream or push back when Simkoff overreaches, and treating those reactions as signals that he is pushing too hard is how he learns to back down.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Max

“without exception, I have gotten each one of them to the point where they were screaming.”

0.52

Doma's title underwriting model uses only property-specific characteristics, not individual borrower or buyer data, to avoid discriminatory or prejudicial underwriting in insurance, while still reaching strong predictive power from hundreds of input variables.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Max Simkoff

“we use only property specific characteristics. So hundreds of input variables, none of which are unique to the individual property transaction.”

0.52

Founders should set aggressive goals, and missing some should not lead to setting less aggressive goals, which is a mistake that reflects kowtowing to normalcy; the goal should be changed only when it is truly unreachable, since abandoning goals early damages morale.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Max Simkoff

“If you miss some of those goals along the way, it does not mean you should set less aggressive goals. That is a mistake.”

0.52

A founder's genius and dysfunction are usually adjacent: Simkoff's dysfunction is that he refuses to let go of disagreements with board members, which led board chair Stuart Miller to yell at him that he was no longer having fun, and a strong board is needed to push back on him.

factualhigh valuespeaker onlynovelty 2/4durability 3/4· Max Simkoff

“I just wouldn't let it go. I refused. I was like, 'I have to do this, whatever.'”

0.52

Growth requires doing both: stay dogmatically excellent in the core title business, since with truly excellent execution it would have about 40% market share and high gross margins, and then expand ambition into adjacencies like appraisal and home warranty once the core is strong.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Max Simkoff

“you get better at your core business. You stay dogmatically committed to being truly excellent at what you're doing.”

0.52

Natural language models can now extract information from free text in deeds at county recorder's offices and identify features relevant to predicting title defect risk, extending data-driven underwriting from binary HR fields to messy unstructured records.

causalhigh valuespeaker onlynovelty 2/4durability 3/4· Max

“now, we can put natural language models against free text in a deed sitting at a county recorder's office”

0.52

Every exceptional founder's dysfunction sits right next to their genius, so the useful practice is to identify your own paired dysfunction and manage it deliberately while leveraging the genius.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Patrick

“anyone with genius, their dysfunction always sits right next to their genius.”

0.52

Doma works like an API for lenders: they send closing requests and receive faster, cheaper results, with machine learning and software replacing the roughly 80 cents on the dollar of cost that front- and back-office labor consumed at legacy title firms.

causalhigh valuespeaker onlynovelty 2/4durability 3/4· Max Simkoff

“Absolutely correct.”

0.52

At Evolv, applying statistics and econometrics to structured HR data produced dramatically better hiring outcomes in hourly workforce selection, showing that a function assumed to be intuitive can be improved with data.

causalhigh valuespeaker onlynovelty 2/4durability 3/4· Max Simkoff

“applying statistics or econometrics against reasonably structured data could yield dramatically better business outcomes in a function where people thought you can't apply data to that, that's got to be an intuitive experiential thing.”

0.52

After closing a Series C on March 13, 2020, Doma's real estate revenue dropped off a cliff during lockdown, forcing layoffs and a reprioritized spending plan within a week or two, because a mistake that required a second cut within six to eight weeks would be a disaster.

factualhigh valuespeaker onlynovelty 2/4durability 3/4· Max Simkoff

“It was on March 13th that we closed our series C financing... The country went into lockdown and the real estate market dropped off a cliff.”

0.52

The hardest part of managing stakeholder perception is retaining long-term conviction while visibly listening to people who challenge your assumptions, because if their point is valid, the future you have promised may be false.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Max

“retaining confidence in the long-term while also allowing the people who justifiably are trying to challenge certain assumptions that you have to show them that you're listening.”

0.50

Simkoff reviews a weekly customer health dashboard with a composite score, sentiment and SLA metrics measured by the minute, and deliberately looks at the bottom outlier customers rather than the top accounts, asking why they are not perfect.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Max Simkoff

“I don't look at the top of the list, I look at bottom.”

0.50

The most important data science hire is someone willing to borrow the best available academic research and experiment aggressively, ignoring the constraints other practitioners accept, which Simkoff argues is why Doma's chief data scientist was hired from astrophysics.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Max

“I would argue that our chief data science officer hire was certainly one of the most important in the company's history”

0.50

When someone says a goal cannot be done, a leader should not immediately cut them off with a reflex that it can be done, since the person may be pointing to missing sales-deck content or product features that are vital information.

normativehigh valuespeaker onlynovelty 2/4durability 3/4· Max Simkoff

“you might miss some other important piece of information that's leading them to say it can't be done.”

0.49

Doma partnered with Lennar, buying Lennar's title business in early 2019, and a reinsurer became a strategic investor, which together gave Doma a large data set, scale, and a cash flow stream that helped it become a going concern.

factualhigh valuespeaker onlynovelty 2/4durability 2/4· Max Simkoff

“We partnered with Lennar. We bought their title business from them in early 2019. And it brought a massive amount of data with it”

0.49

Simkoff's founder 'loose screw' is refusing to accept that something cannot be achieved on a reasonably short timeframe, citing the rapid COVID vaccine development within 12 months as evidence that apparently impossible goals are achievable with purpose and clarity of vision.

normativehigh valuespeaker onlynovelty 2/4durability 2/4· Max Simkoff

“My loose screw is that I refuse to take for an answer that something is not achievable on a reasonably short timeframe.”

0.44

A typical high-end purchase transaction carries total title and escrow fees of roughly $2,500 to $3,000, a retail refinance on a $300-400K loan runs about $1,200, and Doma's bulk enterprise pricing is around $700 total for premium and escrow.

factualhigh valuespeaker onlynovelty 2/4durability 2/4· Max Simkoff

“The high end, a typical purchase transaction, it's probably total title and escrow fees of 2,500 to 3,000 bucks.”

0.43

When hiring quickly, a talented candidate must also be assessed for how they will interact with teammates and other teams, since people want to join and stay where they like the people they will spend time with.

normativeestablishednovelty 1/4durability 3/4· Max Simkoff

“Are people going to look up to this person or are they going to think they're a jerk?”

0.42

Doma's escrow inputs are often flat image PDFs, so it applies OCR to extract text, then natural language models trained to understand the syntax title professionals use, such as how vesting language implies tasks.

factualestablishednovelty 2/4durability 3/4· Max Simkoff

“Many times we have to use optical character recognition just because sometimes the PDFs aren't even structured.”

0.39

The first title companies were effectively groups of lawyers who examined county records for each residential transaction to confirm no outstanding ownership or trust claims conferred to the buyer, a process that changed little in its basic logic until recently.

factualestablishednovelty 1/4durability 3/4· Max Simkoff

“The first title companies were effectively groups of lawyers who would, when a residential real estate transaction was happening, they'd examine the records”

0.34

The real estate industry broadly has historically underinvested in technology, which is why executives from major tech firms rarely come from real estate companies, and this gap is the opportunity Doma is targeting.

factualspeaker onlynovelty 2/4durability 3/4· Max Simkoff

“it's an industry that is historically under invested in technology. You don't hear about like the new CXO at Microsoft coming from a real estate company.”

0.31

Doma identified appraisal and home warranty as problematic, non-digital adjacencies on either side of title and closing, and plans to own them to create an integrated, instant digital homeownership experience now that public-company capital is available to deploy.

forecastspeaker onlynovelty 2/4durability 2/4· Max

“we've identified two adjacencies in the form of appraisal and home warranty”

0.31

Experienced title examiners at incumbent firms do not check statistical signals like zip-code housing appreciation when reviewing a report, so their process leaves predictive information unused even after a century of practice.

causalspeaker onlynovelty 2/4durability 2/4· Max Simkoff

“There's probably not a single title examiner at any one of the other incumbent title companies that have been doing this for 100 years who ever... says, 'I'm just going to look at one more thing'”

0.30

A winning team needs members who feel they are solving problems never solved before, who enjoy the challenge, and who want to spend time with colleagues, and recruiting must balance talent with fit for how a hire will interact with others.

normativespeaker onlynovelty 1/4durability 3/4· Max Simkoff

“It needs to be a place where they feel like they're solving problems they've never been able to solve before.”

0.28

Simkoff proposed to his own bank that he pay $50 and sign a certified affidavit under penalty of perjury attesting to no new ownership interest, in lieu of paying $1,500 again for title insurance, and asked to be charged only if he was wrong.

normativespeaker onlynovelty 2/4durability 2/4· Max Simkoff

“How about I pay you $50 and you send me a certified affidavit”

0.28

The realtor sales channel sits between B2B2C and pure B2C because realtors are independent entrepreneurs who must be persuaded that a solution helps their own business by closing faster with more certainty.

factualspeaker onlynovelty 2/4durability 2/4· Max Simkoff

“a lot of realtors are much more like individuals than they are... They're entrepreneurs.”

0.28

The kindest thing anyone has done for Simkoff was his wife supporting the family while early company income was non-existent or low, while she built her own successful business and shouldered the stress so he could build Doma.

factualspeaker onlynovelty 1/4durability 4/4· Max Simkoff

“My wife supported me to build this company. And when I say support, I mean early days income was either non-existent or low in startup days.”

0.28

A founder challenging a large incumbent should expect to run through consecutively harder brick walls, and an absence of such obstacles suggests the company is not doing anything interesting.

normativespeaker onlynovelty 2/4durability 2/4· Patrick

“As a founder of a company challenging a big incumbent, if you're not running through a consecutively harder brick wall, you're probably not doing anything interesting.”

0.28

Around February or March, a leader should only re-forecast if the year-end goals are truly unreachable; if they are still reachable, the leader should act as cheerleader and will the team to hit them, because second-guessing undermines everything.

normativespeaker onlynovelty 2/4durability 2/4· Max Simkoff

“I think your job is to be like, 'Is there still an ability to hit these goals for the end of the year?' If so, no, we're not doing a re forecast.”

0.26

At the time of Evolv, entry-level hourly workers made up two-thirds of the national workforce, and large employers hired tens of thousands of people per job category each year, creating large structured datasets.

factualspeaker onlynovelty 2/4durability 2/4· Max Simkoff

“which happened to make-up... At the time, it was two thirds of the nation's workforce.”

0.26

In hindsight, Evolv's early models were rudimentary, built with off-the-shelf statistical software like SPSS, and the engineering team had to implement the survey and model outputs in production, with computing costs making it harder.

factualspeaker onlynovelty 2/4durability 2/4· Max Simkoff

“Now in hindsight, it was all very rudimentary.”

0.26

Simkoff agreed that a founder challenging an incumbent should keep running into harder walls, since he finds pleasure in these obstacles and believes a rules-based obstacle like the credit-rating catch-22 is the kind of problem worth solving.

normativespeaker onlynovelty 2/4durability 2/4· Max Simkoff

“That gets back to the screw loose thing. That's probably another screw loose that I have, which is I get some sick pleasure out of this stuff.”

0.21

Doma built a technology platform that uses machine learning to remove most of the friction, frustration, time and expense of closing a mortgage, structured in the footprint of a title company.

definitionspeaker onlynovelty 1/4durability 2/4· Max Simkoff

“Doma has built a technology platform that at it's core uses machine learning to remove most of the friction, frustration, time and now expense involved with closing a mortgage.”

0.21

Hiring Andy Mahdavi, a former tenured astrophysics professor, as chief data science officer was among the most important hires in Doma's history because the company's business is built around its data science.

normativespeaker onlynovelty 1/4durability 2/4· Max Simkoff

“I would argue that our chief data science officer hire was certainly one of the most important in the company's history”