Author: Thomas Crown

When you’re raising venture capital, it helps if you’ve had “exits.” In other words, if your company has been acquired or you’ve taken one public, investors are more inclined to take a bet on anything you do.

Boston -based serial entrepreneur David Cancel has sold not just one, but four companies.  And after a few years running product for HubSpot, he’s in the midst of building number five.

That startup, Drift, managed to raise $47 million in its first three years. Now it’s announcing another $60 million led by Sequoia Capital, with participation from existing investors CRV and General Catalyst. The valuation is undisclosed.

So what is Drift? It’s “changing the way businesses buy from businesses,” said Cancel. He wants to eventually build an alternative to Amazon to make it easier for companies to make large orders.

Currently, Drift subscribers can use chatbots to help turn web visits into sales. It has 100,000 clients including Zenefits, MongoDB, Zuora and AdRoll.

Drift “turns those conversations into customers,” Cancel explained. He said that technology is comparable to what is commonly used for customer service. It’s the “same messaging that was used for support, but used in the sales context.”

In the long-run, Cancel says he hopes Drift will expand its offerings to compete with Salesforce.

The company wouldn’t disclose revenue, but says it is ten times better compared to whatever it was in the past year. And it’s on track to grow another five times this year. This, of course, means little without hard numbers.

Yet we’re told that the new round means that Drift will have $90 million in the bank. It plans to use some of the funding to make acquisitions in voice and video technology. Drift also plans to expand its teams in both Boston and San Francisco, with new offices for both. The company presently has 130 employees.

from TechCrunch

Can Artificial Intelligence replace lawyers? Perhaps sometime in the distant future, but in the meantime AI is already augmenting the work done by legal professionals as startups race to reach that ultimate goal.

One burgeoning player in the AI-powered legal tech space is Tel Aviv-based LawGeex, which has developed automated contract review technology to help companies sift through things like NDAs, supply agreements, purchase orders, and SaaS licenses, to ensure they’re aren’t any unsanctioned legal gotchas buried deep in legalise. Today, the company is announcing that it has closed $12 million in new investment.

Led by VC fund Aleph, with participation from previous backers, including Lool Ventures, the new round of funding will be used by LawGeex to further develop its product, and build a bigger presence in the U.S. where it recently opened a New York office. It brings the startup’s total funding to date to $21.5 million.

Designed to answer the question ‘Can I sign this?’ the LawGeex contract review system aims to significantly speed up and cut costs inherent with the contract approval process. The idea is that once a new contract is sent to a business, it is uploaded to LawGeex where a “first-pass review” of the contract is undertaken using the startup’s AI. This checks the contract against a company’s predefined legal policies.

“If everything looks good, we can automatically approve the contract for signing right then and there,” explains LawGeex VP Marketing Shmuli Goldberg. “If we spot any issues that need to be corrected, we escalate the contract to the legal team, and highlight the exact sentence they need to fix, and what they need to do to fix it”.

The desired outcome is that legal professionals no longer need to spend time reviewing problem-free contracts, and only spend a few minutes, instead of hours, on problematic ones. “We free up the time of whoever does that first review of the contract, be it a paralegal who takes a first look before sending issues on to a lawyer, or a contract review team who triage incoming contracts,” Goldberg says.

Put more simply, the LawGeex product operates a little like a spelling or grammar checker (see screenshot above). But instead of looking for specific keywords or language, the AI has been trained to understand technical legal language or so-called legalese. “It actively reads the contracts and “understands” the legal concepts. This means we can find and flag provisions even if they’re written in a way we’ve never seen before,” says the LawGeex VP.

To make all of this possible, over the last four years the company’s “recursive neural network”-based AI has been trained by feeding it hundreds of thousands of legal contracts, and having experienced U.S. lawyers annotate those contracts along the way. “We’ve now reached the point we can say that in certain cases, for example reviewing standard NDAs, our AI is actually more accurate than a human, as a recent study led by several academics at leading universities showed,” claims Goldberg.

from TechCrunch

Harbor helps businesses legally issue cryptocurrency tokens that represent ownership of real-world assets like real estate, fine art, company equity, and investment funds. This “tokenization” might sound boring, but it could be a big business that unlocks trading of illiquid property.

Harbor‘s intention to become a fundamental bridge between the offline and crypto economies has attracted a $28 million strategic round led by Founders Fund and joined by Andreessen Horowitz, Pantera Capital, and more. Following its $10 million Series A in February, Harbor has now raised over $40 million to dissolve the legal barriers to private securities tokenization.

“We think there’s going be a far greater appetite for owning real-world assets using the blockchain” than digital only cryptocurrencies, Harbor CEO Joshua Stein tells me. He expects it be like the impact “email had on snail mail”, but with value instead of content being sent back and forth. Once someone like Harbor handles the technical necessities to make transfers instant, free, and secure, people will exchange a lot more frequently.

The Harbor team

Here’s how Harbor works. Clients pay it in cash to make their tokenization of an IRL private security legal. Traditional trading of these assets can be complicated and expensive given there are often financial regulations or licensing requirements restricting who can buy and sell them. For example, foreigners or unaccredited investors without enough net worth aren’t allowed to own certain securities. The lawyers to handle these sales can be expensive, and the process can take weeks.

Normally, businesses have to be very careful about who they let buy these securities because they’re liable for a 20-year criminal sentence if they violate SEC law. With Harbor, a white list of eligible owners is established by an outside law firm that takes responsibility, and Harbor’s smart contracts refuse to process an illegal sale. Harbor effectively bakes securities law compliance like know-your-customer and anti-fraud/money-laundering into the tokens themselves so trades can happen instantaneously without legal assistance on every sale.

Harbor is hoping to launch this Regulated Token (R-Token) system with its first client this summer. The tokens are ERC-20 compatible so they can be sold on lots of cryptocurrency exchanges and stored in popular wallets. Stein stresses that investors will have to trust the underlying securities they’re buying. But they’ll get more trust in who owns something through blockchain transparency rather than some signed contract locked in a desk or vault somewhere. And they won’t have to trust who they’re selling to since the smart contracts only execute the trade if its legal.

The idea of making the way hugely valuable assets trade faster, easier, and cheaper led Harbor’s latest round to be oversubscribed. That’s even though it only came out of stealth two months ago from Craft Ventures, the fund and incubator run by PayPal mafioso David Sacks who sold Yammer to Microsoft.

Craft Ventures, Vy Capital and Valor Equity Partners joined this that included other new investors like Future Perfect Ventures, 1confirmation, Abstract Ventures, and Signia Venture Partners. Nicolas Berggruen of Berggruen Holdings, Napoleon Ta of Founders Fund, and Kyle Samani and Tushar Jain of Multicoin Capital also put in their personal money.  Sacks knew Ta, which set up Founders Fund to lead the round. Meanwhile, Stein says Harbor wanted to team up with Andreessen Horowitz partner and crypto thought leader Chris Dixon.

Harbor will have to compete with the other blockchain-for-securities startups like Polymath, which runs entirely decentralized and trustless infrastrucutre to the point that you have to hope strangers want their deposit back enough not to screw you on legal compliance, and tZERO, which is building its own full-stack compliance system. Harbor’s reliance on outside legal firms to build the smart contract white lists makes it more akin to a traditional financial player.

Harbor could make a lucrative business out of letting clients sell American securities to the Chinese market, which has shown a strong interest in crypto assets. Stein talks about “a crypto nirvana of a trustless environment” like a true Bitcoin bro. But his new A-list investors show Harbor is no pump-and-dump.

from TechCrunch

Lingumi, the London and Cardiff-based edtech startup that teaches English to kids aged between 2 and 6 using an app and a range of physical products, has picked up £1.2 million in seed funding.

Leading the round is ADV, with participation from existing backers LocalGlobe, and company builder Entrepreneur First (Lingumi was part of EF’s 5th cohort, which is turning out to be quite a vintage year). A number of unnamed angel investors also took part in the round.

Founded by CEO Toby Mather and CTO Adit Trivedi after they paired up at EF in late 2015, Lingumi has built a language learning platform for pre-school kids, initially targeting the teaching of English. Described as “digital-first,” it consists of an app designed around a curriculum of daily lessons, which can then be augmented with Lingumi’s physical products, such as ‘Play Cubes’ and ‘Jumbo Word Cards’.

In a call with Mather he told me that the Lingumi product that exists today is very different to the one he and Trivedi originally pitched at EF Demo Day back in early 2016, even if the mission remains the same: to increase access to learning a second language, based on a belief that a child’s earliest years present a “magic window” to do so.

Initially, the pair had developed a concept for a connected toy that controlled a learning app but pivoted to a subscription model after families kept hitting the end of the startup’s curriculum and requested more. In July last year, Lingumi ditched the connected toy entirely and switched to a “pure digital subscription,” up selling its now much simpler physical products separately.

“We’re increasingly aware of the exceptional ability of infants to learn a second language from very early in childhood, but access to English is typically restricted to the super rich, or to children aged 7 or older, as they begin to attend school,” says Mather. “Even there, they are taught badly and infrequently. We’re building an English learning methodology that is low-cost, can be used in the earliest years, and is effective, even if the parents themselves don’t speak English”.

The Lingumi platform works best when parents or caregivers participate, too. The app delivers a single lesson per day of around 20 minutes and purposefully limits screen time, hence the range of supplementary non-digital prodicts. “Children receive everything in English, through a playful learning programme built for pre-schoolers, but we encourage and train parents in their native language to play and learn with their children,” explains the Lingumi CEO.

“Multiple studies have shown the impact of this style of co-learning on outcomes. The learning method is also unique: unlike most curriculums, which focus around ‘edutainment’ or reading and writing skills, or teach older children via live video, ours is focused on constructive, natural spoken English in the earliest years. As we develop the curriculum, we’re continually leveraging our data on each child to improve the experience and learning trajectory for them”.

To date, Lingumi claims 10,000 users and Mather says typical customers are families with one or two working parents, “usually middle or working-class, aspirational families who understand both the fun, and the educational benefit of beginning a second language with their child”. It has customers in over 40 countries, but is mainly focussed on Western Europe, Taiwan, and, increasingly, China.

In fact, Mather says China is a potentially huge market and is in part seeing the startup pilot a version of Lingumi for kindergartens that want to begin teaching English, leveraging the company’s existing learning method and content.

To that end, the company plans to use the new seed funding to further develop its “digital and physical product ecosystem,” and scale the learning platform into new markets.

from TechCrunch