Earlybird VC closes new €350M fund for Western European startups, with a deeptech angle

Earlybird VC was one of the, if not the “OG” of the emergence of Berlin as a key global tech startup ecosystem 15 years ago. Founded in 1997, it’s gone on to back N26 and UiPath, among many others. And in the past four months, two more Earlybird portfolio companies have become Unicorns: Aiven, a Finnish software company that combines open source with cloud infrastructure and OneFootball, a German sports and soccer media platform.

It’s now closed its seventh early-stage fund at a hard cap of €350 million, making it one of the largest European early stage funds, and the firm says it was oversubscribed. A couple of years ago Earlybird split into Earlybird West and East, with the latter taking on regions like Central Europe and Turkey. This new fund is anchored in the Digital West (as in Western Europe) investment team.

The Earlybird Digital West fund VII will look at Enterprise Software, Fintech and Sustainability, with a particular focus on deep tech. Earlybird Digital West has made over 17 investments out of this new fund and they include both existing and new portfolio companies such as Aleph Alpha, Deed, Finmid, Hive Technologies, HiveMQ, Marvel Fusion, MAYD, Remberg, Sikoia and ThingsTHINKING.

Hendrik Brandis, Partner and Co-Founder at Earlybird said in a statement: “Our portfolio companies Isar Aerospace, Aleph Alpha, Marvel Fusion or SimScale show that deep tech startups are on the rise and stem from continuous work of scientific institutions across Europe. Our role is to offer these highly scientific young companies, besides our other focus sectors such as Fintech, Enterprise Software and Sustainability, commercialization and growth opportunities on a global scale, in order to make ground-breaking ideas available to society.” 

Christian Nagel, Partner & Co-Founder at Earlybird added: “We are grateful for the high-level of commitment and trust coming from our long-term investors, many of whom accompanied us along almost all our fund generations.”

Earlybird has had some other recent wins: the recent $900 million funding round for N26 – making its the second most valuable retail bank in Germany; and the rise of Isar Aerospaceone of the best-funded space tech company in Europe.

E-commerce logistics startup Hive buzzes following $34M investment

Berlin-based Hive announced $34 million in new investment Wednesday as it continues developing its technology that provides direct-to-consumer brands an alternative way to manage operations.

The company was founded by Oskar Ziegler, Franz Purucker and Leo von Kleist in July 2020 to offer software and operational services, including transparent cost structures, special features and sustainable packaging and shipping.


Hive founders, from left, Oskar Ziegler, Franz Purucker and Leo von Kleist. Image Credits: Hive

Ziegler, who started out in food delivery with foodpanda in Hong Kong, said he began consulting on e-commerce and logistics, and during a discussion with friends who were building DTC brands, the topic turned to operational problems.

“It was then that I thought we could build something to give small retailers the same logistics power as Amazon,” he added.

Purucker explained that when DTC brands get started, the founders are passionate about sales, marketing and finding the right message and influence to get their brand out into the world. However, when they go beyond 30 orders per month to more like 1,000, it becomes more difficult to check inventory from their homes, and they are not getting the cost-competitive rates from freight forwarders that the big brands are getting.

Hive levels that playing field by using data to streamline fulfillment, which he considered “the heart of the operation,” by providing a variety of delivery choices and the ability to get the same inventory management, cost structure and customer notifications that companies like Amazon are using.

The new funding round was led by existing investors Earlybird and Picus and also includes new investor Tiger Global Management. It gives Hive total funding of $44 million and boosts its valuation to $157 million.

“We have looked extensively at the fulfillment space and gained conviction that the full stack model is superior due to its full control over the ultimate service quality,” Fabian Heilemann, partner at Earlybird, said via email. “Hive has compiled the strongest team in the European market to execute on that vision. That’s why we doubled down on our investment in this round.”

Now armed with the new capital, Hive intends to invest in product development and build out new services for supply chain and business teams that will have it triple its tech team by next year. It will also expand geographically, starting with Paris in early 2022. The company aims to bring down delivery times to two days in most places in Europe and next-day for core markets.

In 18 months, Hive quickly grew into a team of more than 100, working on the warehousing of products and the processing of orders through its own fulfillment center. During that time, the company’s revenue grew 10 times and it is processing a few thousand packages per day.

European Investment Fund puts $30M in Fabric Ventures’ new $120M digital assets fund

Despite their rich engineering talent, Blockchain entrepreneurs in the EU often struggle to find backing due to the dearth of large funds and investment expertise in the space. But a big move takes place at an EU level today, as the European Investment Fund makes a significant investment into a blockchain and digital assets venture fund.

Fabric Ventures, a Luxembourg-based VC billed as backing the “Open Economy” has closed $120 million for its 2021 fund, $30 million of which is coming from the European Investment Fund (EIF). Other backers of the new fund include 33 founders, partners, and executives from Ethereum, (Transfer)Wise, PayPal, Square, Google, PayU, Ledger, Raisin, Ebury, PPRO, NEAR, Felix Capital, LocalGlobe, Earlybird, Accelerator Ventures, Aztec Protocol, Raisin, Aragon, Orchid, MySQL, Verifone, OpenOcean, Claret Capital, and more. 

This makes it the first EIF-backed fund mandated to invest in digital assets and blockchain technology.

EIF Chief Executive Alain Godard said:  “We are very pleased to be partnering with Fabric Ventures to bring to the European market this fund specializing in Blockchain technologies… This partnership seeks to address the need [in Europe] and unlock financing opportunities for entrepreneurs active in the field of blockchain technologies – a field of particular strategic importance for the EU and our competitiveness on the global stage.”

The subtext here is that the EIF wants some exposure to these new, decentralized platforms, potentially as a bulwark against the centralized platforms coming out of the US and China.

And yes, while the price of Bitcoin has yo-yo’d, there is now $100 billion invested in the decentralized finance sector and $1.5 billion market in the NFT market. This technology is going nowhere.

Fabric hasn’t just come from nowhere, either. Various Fabric Ventures team members have been involved in Orchestream, the Honeycomb Project at Sun Microsystems, Tideway, RPX, Automic, Yoyo Wallet, and Orchid.

Richard Muirhead is Managing Partner, and is joined by partners Max Mersch and Anil Hansjee. Hansjee becomes General Partner after leaving PayPal’s Venture Fund, which he led for EMEA. The team has experience in token design, market infrastructure, and community governance.

The same team started the Firestartr fund in 2012, backing Tray.io, Verse, Railsbank, Wagestream, Bitstamp, and others.

Muirhead said: “It is now well acknowledged that there is a need for a web that is user-owned and, consequently, more human-centric. There are astonishing people crafting this digital fabric for the benefit of all. We are excited to support those people with our latest fund.”

On a call with TechCrunch Muirhead added: “The thing to note here is that there’s a recognition at European Commission level, that this area is one of geopolitical significance for the EU bloc. On the one hand, you have the ‘wild west’ approach of North America, and, arguably, on the other is the surveillance state of the Chinese Communist Party.”

He said: “The European Commission, I think, believes that there is a third way for the individual, and to use this new wave of technology for the individual. Also for businesses. So we can have networks and marketplaces of individuals sharing their data for their own benefit, and businesses in supply chains sharing data for their own mutual benefits. So that’s the driving view.”

EarlyBird’s new app lets families and friends ‘gift’ investments to children

A new fintech startup called EarlyBird wants to help families invest in their children’s financial futures. Through the EarlyBird mobile app, parents in just a few minutes can create a custodial account, also known as a UGMA (Uniform Gifts to Minors Act) account. These accounts typically allow a parent, aka the “custodian,” to invest in stocks, bonds, mutual funds, and other securities on behalf of the minor child. When the child comes of legal adult age, the investments become theirs.

Through the app, parents can set up an account for their child, then invite other family members and close friends to contribute.

The idea is not so different, in spirit at least, from something like HoneyFund, where newlyweds ask loved ones for cash donations instead of physical gifts. Similarly, EarlyBird offers an alternative to giving a child toys and more “stuff,” by inviting family and friends to donate money. Except in EarlyBird’s case, it’s not asking for straight cash donations — this is not some glorified crowdfunding platform, after all — it’s enabling investments.

Specifically, EarlyBird aims to make it easier and less confusing for parents to establish custodial accounts. It’s not the first fintech to do so — Stash and Acorns, for example, also offer this.

EarlyBird, however, aims to combine the investment account itself with a platform that allows for social features and a gifting experience. The idea is to make the act of donating to the account feel more like a real gift — unlike the gift of a check or some cash tucked into a greeting card.

Image Credits: EarlyBird

With the EarlyBird app, the giver can record a short video “memory” alongside their donation to the investment account. This makes for a more social and personal experience as the child can later look back on these videos. In addition, other family members and friends may also see the videos and be prompted to donate to the child’s investment account, too.

The idea for EarlyBird comes from former AgilityIO COO Jordan Wexler, now EarlyBird CEO, and early Yello.co employee and VP Caleb Frankel, now EarlyBird COO.

Wexler explains that he began thinking about investments as an alternative to physical gifts when a new baby arrived in his own extended family.

“This all started with a problem I experienced years ago when my beautiful baby niece was born. I found myself head over heels and spending hundreds and hundreds of dollars on just the most ridiculous stuff — pretty much just junk gifts,” he says.

A few years ago, he got the idea to start investing his cash into an index fund on the child’s behalf.

“I wanted to have a larger impact in her life and something that she could really use when she grew up,” Wexler says.

His father had once done the same for him, in fact. When he was 12 years old, his dad gave him some money in a TD Ameritrade account which he withdrew later in life to help fund his first startup — SucceedOverseas in Qingdao, China — a strategic consulting firm that aided companies with employee relocation. (It was acquired in 2015 by Chiway Education Group.) 

Wexler met EarlyBird co-founder Caleb Frankel in Qingdao and reconnected with him again when he returned the U.S. Last year, they teamed up on EarlyBird, with the goal of simplifying the process for parents who want to launch custodial investment accounts for their kids.

Image Credits: EarlyBird

Custodial accounts, to be fair, are perhaps not a well-known investment vehicle to those who aren’t parents — or even to those who are, in some cases. That’s because their alternative, the 529 plan, has generally been more popular because of its tax advantages.  

While both accounts allow families to invest on behalf of minor children, investments in 529 plans grow tax-free. Any withdrawals made for educational expenses — like tuition, room and board, books, and more — are also not taxed. That’s a big perk.

UGMA accounts, meanwhile, are taxed at certain levels. The first $1,100 of unearned annual income is tax-free, but the next $1,100 is taxed at the child’s tax rate. Unearned income above $2,200 is then taxed at the rates for trusts and estates, which can be higher than the child’s tax rate.

Donations to UGMA accounts don’t receive an income tax reduction, but they aren’t taxed themselves up to $15K for an individual or $30K for a married couple.

Because most families are investing with college expenses and tax advantages in mind, 529 plans have been better known. But Wexler says things are changing.

“A lot of parents actually have no idea what education and college will look like in 15 years and want something a little bit more flexible,” he explains.

Plus, UGMA accounts can be used for college, if need be. But if college, say, becomes free in the U.S. one day (!!!), the UGMA account’s investments can be used for anything else. That flexibility is why the account is more attractive to some parents these days — and why other fintechs, like Acorns, are entering this market.

However, EarlyBird will expand into 529 plans within a year, it says. It just didn’t start there.

Image Credits: EarlyBird

Another differentiator between EarlyBird and Acorns or Stash’s custodian plans is how EarlyBird incorporates financial literacy into its product.

From birth to 5 years old, the parent manages the child’s account entirely. But when the child is age 6 to 13, parents can show the app to the child in a special “view only” mode where the child can learn about their investments and watch them grow. At 13 to 18, the child can download the app and, alongside their parents, can begin to interact with it. At age 18 (or 21 in some states), the child takes full custody of the account.

EarlyBird also simplifies the act of investing by offering a range of portfolios from conservative to aggressive. On the conservative side, the portfolio is 100% ETF bond-based while the aggressive portfolio is 100% ETF equity-based. Like Acorns, it offers a fixed portfolio model, but it also offers customized portfolios so you can match your investing to your values — like investing in socially responsible businesses. Users can also automate their investments — small or large — on a recurring basis, if they choose.

Image Credits: EarlyBird

The portfolios were designed and built with a team of expert financial advisors led by EarlyBird advisor Evan List, a 12-year VP at Bernstein Private Wealth Management. The company says the portfolios are integrated with a rebalancing engine on the backend that ensures that each equity position stays within a 10% drift of the target allocation that EarlyBird has set within the selected portfolio. It also reviews all portfolios quarterly and rebalances them, if necessary, similar to other robo-investors.

The startup’s investment accounts are currently held with its partner Apex Clearing Corporation, a third-party SEC registered broker-dealer and member of FINRA and Securities Investor Protection Corporation (SIPC). This arrangement protects the investments up to $500,000 total. In time, EarlyBird aims to transition to a broker-dealer itself.

Currently, EarlyBird generates revenue by way of its $3 per month management fee (and $1 per month for each additional child.)

Over time, it will make money much as many fintechs do. It plans to leverage the trades and transactions with Apex Clearing. And as it transitions to a broker-dealer (when a sizable user base and assets under management are achieved), it may pursue a fully-paid lending program, similar to other brokerages.

These programs aren’t live at this time, to be clear, as the startup is only weeks old.

EarlyBird is backed by $2.4 million in funding, led by Network Ventures, in a round closed in November 2020. Other investors include Chingona Ventures, Bridge Investments, Kairos Angels, Takoma Ventures, Subconscious Ventures and various angels.

The app is a free download on iOS.


Xayn is privacy-safe, personalized mobile web search powered by on-device AIs

As TC readers know, the tricky trade-off of the modern web is privacy for convenience. Online tracking is how this ‘great intimacy robbery’ is pulled off. Mass surveillance of what Internet users are looking at underpins Google’s dominant search engine and Facebook’s social empire, to name two of the highest profile ad-funded business models.

TechCrunch’s own corporate overlord, Verizon, also gathers data from a variety of end points — mobile devices, media properties like this one — to power its own ad targeting business.

Countless others rely on obtaining user data to extract some perceived value. Few if any of these businesses are wholly transparent about how much and what sort of private intelligence they’re amassing — or, indeed, exactly what they’re doing with it. But what if the web didn’t have to be like that?

Berlin-based Xayn wants to change this dynamic — starting with personalized but privacy-safe web search on smartphones.

Today it’s launching a search engine app (on Android and iOS) that offers the convenience of personalized results but without the ‘usual’ shoulder surfing. This is possible because the app runs on-device AI models that learn locally. The promise is no data is ever uploaded (though trained AI models themselves can be).

The team behind the app, which is comprised of 30% PhDs, has been working on the core privacy vs convenience problem for some six years (though the company was only founded in 2017); initially as an academic research project — going on to offer an open source framework for masked federated learning, called XayNet. The Xayn app is based on that framework.

They’ve raised some €9.5 million in early stage funding to date — with investment coming from European VC firm Earlybird; Dominik Schiener (Iota co-founder); and the Swedish authentication and payment services company, Thales AB.

Now they’re moving to commercialize their XayNet technology by applying it within a user-facing search app — aiming for what CEO and co-founder, Dr Leif-Nissen Lundbæk bills as a “Zoom”-style business model, in reference to the ubiquitous videoconferencing tool which has both free and paid users.

This means Xayn’s search is not ad-supported. That’s right; you get zero ads in search results.

Instead, the idea is for the consumer app to act as a showcase for a b2b product powered by the same core AI tech. The pitch to business/public sector customers is speedier corporate/internal search without compromising commercial data privacy.

Lundbæk argues businesses are sorely in need of better search tools to (safely) apply to their own data, saying studies have shown that search in general costs around 18% of working time globally. He also cites a study by one city authority that found staff spent 37% of their time at work searching for documents or other digital content.

“It’s a business model that Google has tried but failed to succeed,” he argues, adding: “We are solving not only a problem that normal people have but also that companies have… For them privacy is not a nice to have; it needs to be there otherwise there is no chance of using anything.”

On the consumer side there will also be some premium add-ons headed for the app — so the plan is for it to be a freemium download.

Swipe to nudge the algorithm

One key thing to note is Xayn’s newly launched web search app gives users a say in whether the content they’re seeing is useful to them (or not).

It does this via a Tinder-style swipe right (or left) mechanic that lets users nudge its personalization algorithm in the right direction — starting with a home screen populated with news content (localized by country) but also extending to the search result pages.

The news-focused homescreen is another notable feature. And it sounds like different types of homescreen feeds may be on the premium cards in future.

Another key feature of the app is the ability to toggle personalized search results on or off entirely — just tap the brain icon at the top right to switch the AI off (or back on). Results without the AI running can’t be swiped, except for bookmarking/sharing.

Elsewhere, the app includes a history page which lists searches from the past seven days (by default). The other options offered are: Today, 30 days, or all history (and a bin button to purge searches).

There’s also a ‘Collections’ feature that lets you create and access folders for bookmarks.

As you scroll through search results you can add an item to a Collection by swiping right and selecting the bookmark icon — which then opens a prompt to choose which one to add it to.

The swipe-y interface feels familiar and intuitive, if slightly laggy to load content in the TestFlight beta version TechCrunch checked out ahead of launch.

Swiping left on a piece of content opens a bright pink color-block stamped with a warning ‘x’. Keep going and you’ll send the item vanishing into the ether, presumably seeing fewer like it in future.

Whereas a swipe right affirms a piece of content is useful. This means it stays in the feed, outlined in Xayn green. (Swiping right also reveals the bookmark option and a share button.)

While there are pro-privacy/non-tracking search engines on the market already — such as US-based DuckDuckGo or France’s Qwant — Xayn argues the user experience of such rivals tends to fall short of what you get with a tracking search engine like Google, i.e. in terms of the relevance of search results and thus time spent searching.

Simply put: You probably have to spend more time ‘DDGing’ or ‘Qwanting’ to get the specific answers you need vs Googling — hence the ‘convenience cost’ associated with safeguarding your privacy when web searching.

Xayn’s contention is there’s a third, smarter way of getting to keep your ‘virtual clothes’ on when searching online. This involves implementing AI models that learn on-device and can be combined in a privacy-safe way so that results can be personalized without putting people’s data at risk.

“Privacy is the very fundament… It means that quite like other privacy solutions we track nothing. Nothing is sent to our servers; we don’t store anything of course; we don’t track anything at all. And of course we make sure that any connection that is there is basically secured and doesn’t allow for any tracking at all,” says Lundbæk, explaining the team’s AI-fuelled, decentralized/edge-computing approach.

On-device reranking

Xayn is drawing on a number of search index sources, including (but not solely) Microsoft’s Bing, per Lundbæk, who described this bit of what it’s doing as “relatively similar” to DuckDuckGo (which has its own web crawling bots).

The big difference is that it’s also applying its own reranking algorithms in order generate privacy-safe personalized search results (whereas DDG uses a contextual ads-based business model — looking at simple signals like location and keyword search to target ads without needing to profile users).

The downside to this sort of approach, according to Lundbæk, is users can get flooded with ads — as a consequence of the simpler targeting meaning the business serves more ads to try to increase chances of a click. And loads of ads in search results obviously doesn’t make for a great search experience.

“We get a lot of results on device level and we do some ad hoc indexing — so we build on the device level and on index — and with this ad hoc index we apply our search algorithms in order to filter them, and only present you what is more relevant and filter out everything else,” says Lundbæk, sketching how Xayn works. “Or basically downgrade it a bit… but we also try to keep it fresh and explore and also bump up things where they might not be super relevant for you but it gives you some guarantees that you won’t end up in some kind of bubble.”

Some of what Xayn’s doing is in the arena of federated learning (FL) — a technology Google has been dabbling in in recent years, including pushing a ‘privacy-safe’ proposal for replacing third party tracking cookies. But Xayn argues the tech giant’s interests, as a data business, simply aren’t aligned with cutting off its own access to the user data pipe (even if it were to switch to applying FL to search).

Whereas its interests — as a small, pro-privacy German startup — are markedly different. Ergo, the privacy-preserving technology it’s spent years building has a credible interest in safeguarding people’s data, is the claim.

“At Google there’s actually [fewer] people working on federate learning than in our team,” notes Lundbæk, adding: “We’ve been criticizing TFF [Google-designed TensorFlow Federated] at lot. It is federated learning but it’s not actually doing any encryption at all — and Google has a lot of backdoors in there.

“You have to understand what does Google actually want to do with that? Google wants to replace [tracking] cookies — but especially they want to replace this kind of bumpy thing of asking for user consent. But of course they still want your data. They don’t want to give you any more privacy here; they want to actually — at the end — get your data even easier. And with purely federated learning you actually don’t have a privacy solution.

“You have to do a lot in order to make it privacy preserving. And pure TFF is certainly not that privacy-preserving. So therefore they will use this kind of tech for all the things that are basically in the way of user experience — which is, for example, cookies but I would be extremely surprised if they used it for search directly. And even if they would do that there is a lot of backdoors in their system so it’s pretty easy to actually acquire the data using TFF. So I would say it’s just a nice workaround for them.”

“Data is basically the fundamental business model of Google,” he adds. “So I’m sure that whatever they do is of course a nice step in the right direction… but I think Google is playing a clever role here of kind of moving a bit but not too much.”

So how, then, does Xayn’s reranking algorithm work?

The app runs four AI models per device, combining encrypted AI models of respective devices asynchronously — with homomorphic encryption — into a collective model. A second step entails this collective model being fed back to individual devices to personalize served content, it says. 

The four AI models running on the device are one for natural language processing; one for grouping interests; one for analyzing domain preferences; and one for computing context.

“The knowledge is kept but the data is basically always staying on your device level,” is how Lundbæk puts it.

“We can simply train a lot of different AI models on your phone and decide whether we, for example, combine some of this knowledge or whether it also stays on your device.”

“We have developed a quite complex solution of four different AI models that work in composition with each other,” he goes on, noting that they work to build up “centers of interest and centers of dislikes” per user — again, based on those swipes — which he says “have to be extremely efficient — they have to be moving, basically, also over time and with your interests”.

The more the user interacts with Xayn, the more precise its personalization engine gets as a result of on-device learning — plus the added layer of users being able to get actively involved by swiping to give like/dislike feedback.

The level of personalization is very individually focused — Lundbæk calls it “hyper personalization” — more so than a tracking search engine like Google, which he notes also compares cross-user patterns to determine which results to serve — something he says Xayn absolutely does not do.

Small data, not big data

“We have to focus entirely on one user so we have a ‘small data’ problem, rather than a big data problem,” says Lundbæk. “So we have to learn extremely fast — only from eight to 20 interactions we have to already understand a lot from you. And the crucial thing is of course if you do such a rapid learning then you have to take even more care about filter bubbles — or what is called filter bubbles. We have to prevent the engine going into some kind of biased direction.”

To avoid this echo chamber/filter bubble type effect, the Xayn team has designed the engine to function in two distinct phases which it switches between: Called ‘exploration’ and (more unfortunately) ‘exploitation’ (i.e. just in the sense that it already knows something about the user so can be pretty certain what it serves will be relevant).

“We have to keep fresh and we have to keep exploring things,” he notes — saying that’s why it developed one of the four AIs (a dynamic contextual multi-armed bandit reinforcement learning algorithm for computing context).

Aside from this app infrastructure being designed natively to protect user privacy, Xayn argues there are a bunch of other advantages — such as being able to derive potentially very clear interests signs from individuals; and avoiding the chilling effect that can result from tracking services creeping users out (to the point people they avoid making certain searches in order to prevent them from influencing future results).

“You as the user can decide whether you want the algorithm to learn — whether you want it to show more of this or less of this — by just simply swiping. So it’s extremely easy, so you can train your system very easily,” he argues.

There is potentially a slight downside to this approach, too, though — assuming the algorithm (when on) does some learning by default (i.e in the absence of any life/dislike signals from the user).

This is because it puts the burden on the user to interact (by swiping their feedback) in order to get the best search results out of Xayn. So that’s an active requirement on users, rather than the typical passive background data mining and profiling web users are used to from tech giants like Google (which is, however, horrible for their privacy).

It means there’s an ‘ongoing’ interaction cost to using the app — or at least getting the most relevant results out of it. You might not, for instance, be advised to let a bunch of organic results just scroll past if they’re really not useful but rather actively signal disinterest on each.

For the app to be the most useful it may ultimately pay to carefully weight each item and provide the AI with a utility verdict. (And in a competitive battle for online convenience every little bit of digital friction isn’t going to help.)

Asked about this specifically, Lundbæk told us: “Without swiping the AI only learns from very weak likes but not from dislikes. So the learning takes place (if you turn the AI on) but it’s very slight and does not have a big effect. These conditions are quite dynamic, so from the experience of liking something after having visited a website, patterns are learned. Also, only 1 of the 4 AI models (the domain learning one) learns from pure clicks; the others don’t.”

Xayn does seem alive to the risk of the swiping mechanic resulting in the app feeling arduous. Lundbæk says the team is looking to add “some kind of gamification aspect” in the future — to flip the mechanism from pure friction to “something fun to do”. Though it remains to be seen what they come up with on that front.

There is also inevitably a bit of lag involved in using Xayn vs Google — by merit of the former having to run on-device AI training (whereas Google merely hoovers your data into its cloud where it’s able to process it at super-speeds using dedicated compute hardware, including bespoke chipsets).

“We have been working for over a year on this and the core focus point was bringing it on the street, showing that it works — and of course it is slower than Google,” Lundbæk concedes.

“Google doesn’t need to do any of these [on-device] processes and Google has developed even its own hardware; they developed TPUs exactly for processing this kind of model,” he goes on. “If you compare this kind of hardware it’s pretty impressive that we were even able to bring [Xayn’s on-device AI processing] even on the phone. However of course it’s slower than Google.”

Lundbæk says the team is working on increasing the speed of Xayn. And anticipates further gains as it focuses more on that type of optimization — trailing a version that’s 40x faster than the current iteration.

“It won’t at the end be 40x faster because we will use this also to analyze even more content — to give you can even broader view — but it will be faster over time,” he adds.

On the accuracy of search results vs Google, he argues the latter’s ‘network effect’ competitive advantage — whereby its search reranking benefits from Google having more users — is not unassailable because of what edge AI can achieve working smartly atop ‘small data’.

Though, again, for now Google remains the search standard to beat.

“Right now we compare ourselves, mostly against Bing and DuckDuckGo and so on. Obviously there we get much better results [than compared to Google] but of course Google is the market leader and is using quite some heavy personalization,” he says, when we ask about benchmarking results vs other search engines.

“But the interesting thing is so far Google is not only using personalization but they also use kind of a network effect. PageRank is very much a network effect where the most users they have the better the results get, because they track how often people click on something and bump this also up.

“The interesting effect there is that right now, through AI technology — like for example what we use — the network effect becomes less and less important. So actually I would say that there isn’t really any network effect anymore if you really want to compete with pure AI technology. So therefore we can get almost as relevant results as Google right now and we surely can also, over time, get even better results or competing results. But we are different.”

In our (brief) tests of the beta app Xayn’s search results didn’t obviously disappoint for simple searches (and would presumably improve with use). Though, again, the slight load lag adds a modicum of friction which was instantly obvious compared to the usual search competition.

Not a deal breaker — just a reminder that performance expectations in search are no cake walk (even if you can promise a cookie-free experience).

An opportunity for competition?

“So far Google has so far had the advantage of a network effect — but this network effect gets less and less dominant and you see already more and more alternatives to Google popping up,” Lundbæk argues, suggesting privacy concerns are creating an opportunity for increased competition in the search space.

“It’s not anymore like Facebook or so where there’s one network where everyone has to be. And I think this is actually a nice situation because competition is always good for technical innovations and for also satisfying different customer needs.”

Of course the biggest challenge for any would-be competitor to Google search — which carves itself a marketshare in Europe in excess of 90% — is how to poach (some of) its users.

Lundbæk says the startup has no plans to splash millions on marketing at this point. Indeed, he says they want to grow usage sustainably, with the aim of evolving the product “step by step” with a “tight community” of early adopters — relying on cross-promotion from others in the pro-privacy tech space, as well as reaching out to relevant influencers.

He also reckons there’s enough mainstream media interest in the privacy topic to generate some uplift.

“I think we have such a relevant topic — especially now,” he says. “Because we want to show also not only for ourselves that you can do this for search but we think we show a real nice example that you can do this for any kind of case.

“You don’t always need the so-called ‘best’ big players from the US which are of course getting all of your data, building up profiles. And then you have these small, cute privacy-preserving solutions which don’t use any of this but then offer a bad user experience. So we want to show that this shouldn’t be the status quo anymore — and you should start to build alternatives that are really build on European values.”

And it’s certainly true EU lawmakers are big on tech sovereignty talk these days, even though European consumers mostly continue to embrace big (US) tech.

Perhaps more pertinently, regional data protection requirements are making it increasing challenging to rely on US-based services for processing data. Compliance with the GDPR data protection framework is another factor businesses need to consider. All of which is driving attention onto ‘privacy-preserving’ technologies.


Xayn’s team is hoping to be able spread its privacy-preserving gospel to general users by growing the b2b side of the business, according to Lundbæk — so it’s hoping some home use will follow once employees get used to convenient private search via their workplaces, in a small-scale reverse of the business consumerization trend that was powered by modern smartphones (and people bringing their own device to work).

“We these kind of strategies I think we can step by step build up in our communities and spread the word — so we think we don’t even need to really spend millions of euros in marketing campaigns to get more and more users,” he adds.

While Xayn’s initial go-to-market push has been focused on getting the mobile apps out, a desktop version is also planned for Q1 next year.

The challenge there is getting the app to work as a browser extension as the team obviously doesn’t want to build its own browser to house Xayn. tl;dr: Competing with Google search is mountain enough to climb, without trying to go after Chrome (and Firefox, and so on).

“We developed our entire AI in Rust which is a safe language. We are very much driven by security here and safety. The nice thing is it can work everywhere — from embedded systems towards mobile systems, and we can compile into web assembly so it runs also as a browser extension in any kind of browser,” he adds. “Except for Internet Explorer of course.”

DeepCode gets $4M to feed its AI-powered code review tool

DeepCode, a Swiss startup that’s using machine learning to automate code reviews, has closed a $4M seed round, led by European VC firm Earlybird, with participation from 3VC and existing investor btov Partners.

The founders described the platform as a sort of ‘Grammarly for coders’ when we chatted to them early last year. At the they were bootstapping. Now they’ve bagged their first venture capital to dial their efforts up.

DeepCode, which is spun-out of Swiss technical university ETH Zurich, says its code review AI is different because it doesn’t just pick up syntax mistakes but is able to determine the intent of the code because it processes millions of commits — giving it an overview that allows it to identify many more critical bugs and vulnerabilities than other tools.

“All of the static analysis and lint tools out there (there are hundreds of those) are providing similar code analysis services but without the deeper understanding of code, and mostly focusing on one language or specific languages,” says CEO and co-founder, Boris Paskalev, going on to name-check the likes of CA Technologies, Micro Focus (Fortify), Cast Software, and SonarSource as the main competitors DeepCode is targeting.

Its bot is free for enterprise teams of up to 30 developers, for open source software, and for educational use.

To use it developers connect DeepCode with their GitHub or Bitbucket accounts, with no configuration required. The bot will then immediately start reviewing each commit — picking up issues “in seconds”.  (You can see a demo of the code review tool here.)

“We do not disclose developer information but the number of Open Source Repositories that are using DeepCode have hundreds of thousands of total contributors,” Paskalev tells us when asked how many developers are using the tool now.

“We do not count rules per se as our AI Platform combines thousands of programming concepts, which if combined in individual rules will result in millions of separate rules,” he adds.

The seed funding will go on supporting additional integrations and more programming languages than the three currently supported (namely: Java, JavaScript, and Python); on improving the scope of code recommendations, and on expanding the team internationally.

Commenting in a statement, Christian Nagel, partner and co-founder of Earlybird, said: “For all industries and almost every business model, the performance and quality of coding has become key. DeepCode provides a platform that enhances the development capabilities of programmers. The team has a deep scientific understanding of code optimization and uses artificial intelligence to deliver the next breakthrough in software development.”