Microsoft announced today that it has acquired Semantic Machines, a Berkeley-based startup that wants to solve one of the biggest challenges in conversational AI: making chatbots sound more human and less like, well, bots.
In a blog post, Microsoft AI & Research chief technology officer David Ku wrote that “with the acquisition of Semantic Machines, we will establish a conversational AI center of excellence in Berkeley to push forward the boundaries of what is possible in language interfaces.”
According to Crunchbase, Semantic Machines was founded in 2014 and raised about $20.9 million in funding from investors including General Catalyst and Bain Capital Ventures.
In a 2016 profile, co-founder and chief scientist Dan Klein told TechCrunch that “today’s dialog technology is mostly orthogonal. You want a conversational system to be contextual so when you interpret a sentence things don’t stand in isolation.” By focusing on memory, Semantic Machines’ AI can produce conversations that not only answer or predict questions more accurately, but also flow naturally.
Instead of building its own consumer products, Semantic Machines focused on enterprise customers. This means it will fit in well with Microsoft’s conversational AI-based products, including Microsoft Cognitive Services and Azure Bot Service, which are used by one million and 300,000 developers, respectively, and virtual assistants Cortana and Xiaolce.
Adzuna, the meta-search engine for jobs, has raised £8 million in Series C funding from Smedvig Capital. The U.K. company’s previous backers include Index Ventures, Passion Capital, LocalGlobe, and over 400 Crowdcube investors. It takes total funding to £12 million.
Founded by the team behind Gumtree, Zoopla and Qype, Adzuna essentially aggregates job listings across the web to offer a single destination to search for a job. It launched first in the U.K. in 2011 but has since expanded to 16 countries, in which co-founder Doug Monro tells me the U.K., U.S., Germany, Netherlands, France, and Brazil are its strongest markets.
“We’re growing very quickly in several of the others. We are really excited about the growth we are seeing in the U.S. in particular,” he says.
Across the 16 sites Adzuna operates, the jobs search engine is seeing 10 million monthly visitors, and has 7 million registered users. “Millions” of CVs have been uploaded to the site — no doubt drawn in by Adzuna’s data-driven “ValueMyCV” tool — and it currently aggregates 5,000 sources of jobs. But, perhaps more importantly, given its Series C backing, the company is disclosing over £1 million in monthly revenue.
Adzuna generates revenue by referring job seekers to jobs. Job ads are included for free in its search index to ensure it always lists every job available, but advertisers have the option to promote listings on a CPC basis similar to Google Adwords. “Some additional revenue is generated through labour market data sales and of course now from the Find a job contract which is publicly disclosed,” says Monro.
The ‘Find a job’ contract is major recent win for the company that saw it displace competitor Monster who ran the pre-existing Universal Jobmatch service for the U.K. government’s Department for Works and Pensions. The publicly procured contract is said to be worth £2.5 million per year.
“We’d been talking to the DWP for a number of years about our vision for how we could help use our tech to help make their service better,” Monro tells me. “Last year they decided to put the Universal Jobmatch out to tender. As a startup with little govtech experience, we thought we had very little chance, but with the help of the Public.io team, we gave it a shot. There was a lot of paperwork and processes to navigate, and we were lucky to have great mentors to help guide us through this, but we were also pleasantly surprised with how agile and open to change the DWP team were”.
Meanwhile, on who Adzuna’s most direct competitors are these days, Monro says there are a number of other job search engines that aggregate content in a similar way but that he believes the startup is taking the market to the next level by bringing innovative tools and smart data to bear, such as the ValueMyCV tool and machine-learning based matching. “It’s a huge market and we are focussed on building the best solution for job seekers. We see ourselves as competing in that sense with the likes of Indeed, Ziprecruiter and LinkedIn,” he says.
Pluralsight is having a pretty good day in its debut as a public company, with its shares popping more than 30% after its first trade following its IPO.
There’ll be a little bit of debate as to whether Pluralsight might have left some money on the table in its IPO after raising its price last night above its original target range. After looking at a range between $12 and $14 per share, the company settled on $15 in an IPO that would raise as much as $357 including additional shares offered to underwriters. But the significant pop this morning suggests that there is both a lot of demand for the company, and also that it could have potentially captured more capital in its IPO.
Still, Pluralsight will be considered a pretty successful one this morning, much like zScalar and Dropbox before it. Pluralsight, like many other enterprise-focused companies, offers investors an opportunity to tap a business model that can grow more consistently and methodically than a consumer company subject to the whims of fickle consumers. While Dropbox has more of a hybrid model, it was considered a substantially successful enterprise IPO, as was zScalar and others earlier this year.
Pluralsight offers companies a way to run courses that help their employees pick up new software engineering skills. That’s important for larger companies that can have a sprawling employee base, offering them an opportunity to find talent in their own workforces that might be missing a few skills instead of having to look out in a very competitive landscape. That minted another successful unicorn startup in the Utah tech scene, and now the company is going to potentially offer a nice return for its investors and an opportunity for investors with some appetite for risky early IPOs .
The company launched in 2004 and was largely bootstrapped until its first financing round in 2013, and raised nearly $200 million total prior to going public. Having a successful IPO like this one is also going to have the ancillary effect of keeping up morale at the company, as well as attracting talent with generous compensation packages. Pluralsight can point to the pop in its IPO and ongoing performance as a public barometer of its success, and the interest Wall Street has in it going forward as a good investment.
It’s been a busy week for startup funding in Southeast Asia. Following big deals for Carro and Carousell, financial comparison startup Jirnexu is the latest to announce new capital after it closed an $11 million Series B round.
The new investment comes courtesy of Japan’s SBI Group — a returning investor which led the round having co-led Carro’s $60 million raise — alongside new backer SIG Asia Investments. The deal takes Jirnexu to $17 million from investors to date.
The startup was founded in 2012 and it is based in Kuala Lumpur, Malaysia. It operates financial comparison services in its native Malaysia (‘RinggitPlus‘) and in Indonesia, under the ‘KreditGoGo‘ brand, that aggregate offerings from banks and financial companies that include Citibank, HSBC, Standard Chartered, and UOB. In short, the company acts as a user acquisition channel for financial organizations that want to reach consumers and maintain a dialogue with them.
In recent years, Jirnexu has gone beyond basic banking products to offer insurance and e-policies, while it has introduced chatbots in conjunction with five financial organizations to help ease the process of sign-up and selection for their customers.
“Our core focus is to become the only band of services a consumer needs for their personal finance and money,” Jirnexu CEO Yuen Tuck Siew, who founded the company after returning home from a decade in the UK, told TechCrunch in an interview. “Two years ago, it was all about banking, particularly secure credit, now we’ve announced live quotes for insurance and we’ll be adding more insurance products.”
In particular, the startup is focused on consumer digital identification and eKYC that will help it to tailor suggested packages more accurately for consumers.
Jirnexu has raised new funding in 2016 and 2017, but Siew said this newest round gives significant runway that will allow it to focus on longer-term strategies with more clarity than before.
“We can now plan multiple years ahead which is absolutely essential with what we can do. No matter how longterm you want to think, [when you need to raise money regularly] you’re always looking at KPIs. Now we can plan and invest in projects that can really have a huge impact for customers,” he explained.
Much of the effort right now, he added, is on hiring for senior executive positions and operational roles, including a CTO, to build out the business and push into new financial verticals.
For now, the company isn’t looking to expand to new markets. Siew suggested that new launches would likely come after a Series C round but, per earlier comments, that’s not an event he sees happening in the immediate term.
Likewise, he said there is potential to work more closely with SBI — which operates a range of financial services in Japan — in the future. But initially, the company is just focused on executing on its plans with its investors’ backing.
“They understand the challenges in the market and see the value of us being able to overcome issues like regulatory and long sales cycles,” the Jirnexu CEO said.
Pluralsight priced the shares in its IPO at $15 this afternoon, above its previously set target range of between $12 and $14, and will raise as much as $357 million ahead of its public debut tomorrow morning.
Pluralsight offers software development courses, specifically ones targeting employees that are looking to advance in their careers by acquiring new skills in order to transition to higher-level roles. As knowledge workers become increasingly valuable, especially in larger enterprises with sprawling workforces, companies like Pluralsight have found a sweet spot in building tools that enable companies to help identify talent in their own workforce and train them, rather than have to aggressively search outside the company to satisfy their needs. The company has raised $310.5 million in its IPO, with underwriters having the option to purchase an additional 3.1 million shares and bring that up to $357 million.
The company is one of a continuing wave of enterprise IPOs this year, including multiple successful ones like zScalar and Dropbox — the latter of which was more of a flagship as both a hotly-anticipated one and as a company that possesses a unique business model. But nonetheless, it’s shown that there’s an appetite for enterprise startups looking to go public, which offers those companies a way to raise capital in addition to offering their employees liquidity.
Pluralsight will be another of an increasing pack of unicorns in the Utah tech scene that are on their way to going public. Founded in 2004, Pluralsight was largely bootstrapped until its first financing round in 2013 where it raised $27.5 million from Insight Venture Partners. That firm is the company’s largest shareholder, and since then Pluralsight has raised nearly $200 million in financing.
Its The company’s IPO tomorrow will once again test the appetite for fresh IPOs among public investors. Enterprise companies generally offer a more stable batch for venture portfolios, with predictable and reliable growth that eventually carries it to an IPO with varying levels of success. They’re smaller than blockbuster consumer-ish IPOs, but they are the ones that can provide a stable return for funds like IVP.
Cryptocurrency startup Circle has raised a $110 million funding round, which values the company near $3 billion. Cryptocurrency mining company Bitmain is leading the round.
Existing investors IDG Capital, Breyer Capital, General Catalyst, Accel, Digital Currency Group and Pantera are investing more money. Blockchain Capital and Tusk Ventures are investing in Circle for the first time. Goldman Sachs also invested in the company in a previous round.
It’s hard to describe Circle in a few words because the company has been active on all fronts. For a really long time, the company pitched itself as a social payment company, a Venmo and Square Cash competitor. But Circle is more focused than ever on cryptocurrencies.
The company has been operating one of the largest over-the-counter trading desk for big cryptocurrency investors and exchanges. Circle Trade manages more than $2 billion a month in transactions and is able to fulfill large orders and provide liquidity.
More recently, the company launched Circle Invest, a really simple mobile app for the U.S. market. It lets you buy and sell Bitcoin, Bitcoin Cash, Ethereum, Ethereum Classic, Litecoin, Zcash and Monero in just a few taps. It’s a good way to get started with cryptocurrencies without learning about exchanges and order types. It could become a good Coinbase competitor for small cryptocurrency investors.
And Circle also acquired Poloniex, one of the largest cryptocurrency exchanges in the U.S.
But the most interesting projects right now are probably CENTRE and a new tokenized USD coin. There are so many different cryptocurrencies, fiat currencies, exchanges and wallets that it has become hard to make everything work together. Cryptocurrencies still suffer from price volatility, so bitcoin can’t be the common denominator.
That’s why Circle is creating a token that is pegged to the U.S. dollar. The USD Coin is based on an open source framework developed by CENTRE and everything should be audited regularly.
CENTRE is a Circle initiative to create a common framework to connect all electronic wallets. This protocol could let you send money to an Alipay user with your Square Cash balance.
It’s clear that Circle wants to build the infrastructure of the cryptocurrency industry. The company will need to convince multiple industry players to work with Circle. But it could help the cryptocurrency ecosystem as a whole.
The HR departments of large companies face a common challenge: how to scale the hiring process when they receive hundreds if not thousands of applicants, and how to remove unconscious bias so the best or most suitable candidates are shortlisted. That’s the specific problem MeVitae, an Oxford-based startup founded by neuroscientist Riham Satti and computer scientist Vivek Doriaiswamy, has set out to solve.
Dubbed “Augmented Intelligence,” more broadly the burgeoning company is developing AI technology that uses what Satti describes as “cognitive techniques” designed to compensate for the limitations of humans.
The premise — in part backed up by her neuroscience research at Oxford University — is that our brains have limitations that restrict our cognitive ability, including the relatively slow speed when processing information, unconscious biases, and limited memory. Limitations, she says, that machines do not possess. Applying to this to various aspects of recruitment is the startup’s initial bet.
Intended to be deployed after a new job opening has been advertised, the MeVitae software plugs into a company’s current Application Tracking System. It then sifts through all of the applications/CVs that have been received and analyses each CV (relative to the job spec) giving it a score.
“This is done by analysing every component of a CV (e.g. education, experience etc.) and using the web to reason and validate each score,” explains Satti.
The hiring company then receives the ranked and shortlisted applicants within their ATS system, and — crucially — is also able to see a “road map” explaining MeVitae’s reasoning behind each score. In addition, through the use of NLP, MeVitae takes each CV and find parts of it that could result in discrimination (e.g. gender, ethnicity) and redacts this information for an employer i.e. CV blinding.
The result is that employers now have ranked and redacted applicant CVs and can quickly shortlist top and diverse talent. “The ranked and anonymised candidates are provided to the recruiter/employer of the company to review,” says the MeVitae co-founder. “The employer will decide who they want to interview from the shortlist… [and] over time the system learns from each employer’s choices for more intelligent decision-making”. In other words, the longer MeVitae is employed by an individual company the more responsive it becomes to that company’s hiring priorities and isn’t a one size fits all solution.
Meanwhile, MeVitae is disclosing that it has raised £500,000 in funding, in a round led by angel investor club Startup Funding Club. Others participating include Force Over Mass, Twenty Ten Capital, BBH ZAG (brand arm of Bartle Bogle Hegartyand), and the tax-payer funded London Co-Investment Fund (launched by the Deputy Mayor of London). Dhiraj Mukherjee, (co-founder of Shazam), Simon Samuel (search and recruitment executive), and Geoff Hughes (Microsoft director and Honorary Research Associate at UCL) have also invested and join the MeVitae board.
Prior to this, the startup has been funded by a number of grants, totalling around £250,000. These came from Innovate UK, European Space Agency BIC, and Regional Growth Fund, amongst others. The funding was used to finance over 3 years of R&D as MeVitae built the technology and got its current product offering to market.
Teatime Games, a new Icelandic “social games” startup from the same team behind the hugely popular QuizUp (acquired in by Glu Mobile), is disclosing $9 million in funding, made up of seed and Series A rounds.
Index Ventures led both, but have been joined by Atomico, the European VC fund founded by Skype’s Niklas Zennström, for the $7.5 million Series A round. I understand this is the first time the two VC firms have done a Series A deal together in over a decade.
Both VCs have a decent track record in gaming. Index counts King, Roblox and Supercell as previous gaming investments, whilst Atomico also backed Supercell, along with Rovio, and most recently Bossa Studios.
As part of the round, Guzman Diaz of Index Ventures, Mattias Ljungman of Atomico, and David Helgason, founder of Unity, have joined the Teatime Games board of directors.
Meanwhile, Teatime Games is keeping shtum publicly on exactly what the stealthy startup is working on, except that it plays broadly in the social and mobile gaming space. In a call with co-founder and CEO Thor Fridriksson yesterday, he said a little more off the record and on condition that I don’t write about it yet.
What he was willing to describe publicly, however, is the general problem the company has set out to solve, which is how to make mobile games more social and personalised. Specifically, in a way that any social features — including communicating with friends and other players in real-time — enhances the gameplay rather than gets in its way or is simply bolted on as an adjunct to the game itself.
The company’s macro thesis is that games have always been inherently social throughout different eras (e.g. card games, board games, arcades, and consoles), and that most games truly come to life “through the interaction between people, opponents, and the audience”. However, in many respects this has been lost in the age of mobile gaming, which can feel like quite a solitary experience. That’s either because they are single player games or turn-based and played against invisible opponents.
Teatime plans to use the newly-disclosed investment to double the size of its team in Iceland, with a particular focus on software engineers, and to further develop its social gaming offering for third party developers. Yes, that’s right, this is clearly a developer platform play, as much as anything else.
On that note, Atomico Partner Mattias Ljungman says the next “breakout opportunity” in games will see a move beyond individual studios and titles to what he describes as fundamental enabling technologies. Linked to this he argues that the next generation of games companies being developed will “become ever more mass market and socially connected”. You can read much more on Ljungman and Atomico’s gaming thesis in a blog post recently published by the VC firm.
French startup Aircall has raised a founding round of $29 million for its cloud based call center solution. Draper Esprit led the round with NextWorld Capital, Balderton Capital and Newfund also participating.
The company has raised $40.5 million in total. Aircall participated in the Startup Battlefield at TechCrunch Disrupt SF a few years ago. The company first started at eFounders.
Aircall is following the software-as-a-service playbook. First, you take a boring industry like phone systems for large support and sales teams. Second, you bet everything on software. And third, you keep adding new features and integrations, and chasing new customers.
The company now has two offices in New York and Paris and handles millions of calls every day. With today’s funding round, the company plans to hire more people in both offices.
When you sign up to Aircall, you get virtual phone numbers in one or multiple countries. You can then configure a greeting message, add business hours and handle your call queue.
But the magic happens when you have multiple people handling sales or support calls. When someone calls, it can call multiple people at once or call someone first, then a second person if the first person isn’t available, etc. You get an overview of all your calls so you can assign them, tag them and more.
Aircall doesn’t work in a vacuum. So you can integrate Aircall with CRMs and other solutions like Salesforce, Zendesk and Zoho. The startup also launched a deep integration with Intercom that lets you switch from a text conversation to a phone call from the popup window.
It’s hard to list all the features right here. But chances are that if you’re running a call center, you’ll have everything you need for your team. Aircall currently costs $30 to $50 per user and per month to access all of this.
Mercari, the eBay-like service that is Japanese first tech startup unicorn, has filed to go public in an IPO that could raise as much as $1.1 billion.
The company is scheduled to list on the Toyko Stock Exchange’s Mothers Market — a board for high-growth companies — on June 19.
The company reached the symbolic $1 billion valuation mark in 2016 when it raised a $75 million Series D. In doing so it became the first Japanese tech startup to become a pre-IPO unicorn. Earlier this year, that valuation jumped to $2 billion following a $47 million investment.
The five-year-old company operates an online ‘flea market’ that lets consumers sell unwanted goods with a focus on mobile.
Japan is its core market, but the company expanded into the U.S. in 2014 and last year it entered Europe, initially via the UK. It boosted its overseas strategy in June 2017 when it hired former Facebook executive John Lagerling as its first chief business officer to guide its global strategy.
The business passed 100 million downloads worldwide at the end of 2017. Mercari said that over 30 million downloads are in the U.S., with more than 60 million in Japan. Speaking earlier this year, CEO Shintaro Yamada — who sold his previous startup Unoh to gaming firm Zynga in 2010 — said success in the U.S. is essential if Mercari is to become an international player.
Reuters reports that Mercari’s forecasted share price of 2,200-2,700 JPY per share would see the company raise up to 117.6 billion JPY ($1.1 billion) at a total market cap of 365.4 billion JPY, $3.3 billion.
It’s common for Japanese startups to go public, but it traditionally tends to happen much earlier than in the U.S or other parts of the world. That’s often times down to investors — who seek to reduce the risk of their money not returning — and a relative lack of capital for startups, but Mercari has held out longer than most and that might set an example for future companies.
For another thing, the return on investment is impressive for many of Mercari’s backers, according to data from 500 Startups partner Yohei Sawayama — who tweeted out USD estimates for potential returns.