Category: UNCATEGORIZED

Instagram today is launching Shopping in Reels, its TikTok competitor. The new feature was announced in October as something the company had in the works, as part of a ongoing series of shopping-related updates to the Instagram app. With today’s launch, both businesses and creators will be able to tag products when they create Reels — the short-form videos that now have their own tab in Instagram following last month’s redesign.

The company says many Reels already feature shopping content, like fashion looks, makeup and skincare, or other product how-tos. When people view an Instagram Reel with this content, they’ll be able to now tap a “View Products” button to either buy, save or learn more about the featured products.

In addition, creators can add a “Branded Content” tag to their Reels to be transparent about when they’re working with a brand to promote their products, which is a form of paid promotion.

The update makes shopping an even larger focus for Instagram than it already was, and arrives at a time when video-based shopping is seeing increased adoption. In particular, a growing number of startups focused on live-stream video shopping are finding traction. In recent months, investors have backed companies like live shopping platforms Popshop Live and Bambuser, for example, while major tech companies, including Alibaba, Amazon, Google and JD.com, have also joined the video shopping trend in various ways.

Most importantly, perhaps, is that Instagram rival TikTok recently partnered with Shopify on e-commerce and today caters to brands that either advertise directly or work with influencers on its platform, eating into Instagram’s market. TikTok had also fielded interest from Walmart, when the Trump ban had forced the company to enter negotiations around a U.S. exit. And TikTok’s app in 2020 beat out Instagram as one of the world’s most downloaded apps, indicating a seismic shift in how younger users interact with social media.

Hoping to not be left behind, Instagram has revamped its app — to much user criticism when it relocated key home screen features — with the goal of becoming a top online shopping destination, as well. The company generates revenue when customers checkout in the app using Facebook Pay, which will allow it to make money outside of running ads.

Today, Instagram users can shop from videos in Feed, Stories, Live, IGTV and, with this latest launch, Reels.

The company says the feature is rolling out globally, starting today.

Cleo, the London founded “financial assistant” that takes the form of an app and chatbot and now counts the U.S. as its largest market, has raised $44 million in Series B funding.

Leading the round, which I understand actually closed earlier this year, is EQT Ventures. Also participating are existing investors Balderton Capital, LocalGlobe and SBI.

They join much earlier investors such as Entrepreneur First, Taavet Hinrikus, Matt Robinson, Errol Damelin, Niklas Zennstrom, Alex Chesterman, and Ian Hogarth — all well-known names in London’s tech investment community.

Targeting “Gen Z” and with a rather lofty sounding mission to “fight for the world’s financial health,” Cleo’s AI/machine learning-powered app connects to your bank accounts and gives you proactive advice and information on your finances, including timely nudges, to help you stay on top of your spending. Over time, the idea is that Cleo can help change your financial behaviour for the better.

The broader premise is that Cleo can replace your bank’s own app, and by speaking to you in a more human and user-friendly way, improve your financial health. In addition, and crucially — as Cleo founder Barney Hussey-Yeo is fond of arguing — the company can do all of this without having the same cost base or misaligned incentives of an actual bank.

The assumption made in 2017, when Cleo really got off the ground, was that through a combination of data science and machine learning, delivered in a fun and openly gamified way, the company’s financial assistant chatbot could attract and retain gen z and millennial users. Then, after engaging with the app, users would have experienced enough value to upgrade to a paid subscription or various financial services offered through Cleo now or in the future. With 4 million registered users — 96 percent of whom are in the U.S. — the first part appears to be panning out.

Hussey-Yeo tells me the new funding round gives him a “mandate” to take some big product bets in order to move forward the financial health of Cleo’s users. “[Those product bets] won’t all work out, but if one or two of them connects, we’ll fundamentally change the game in some of the most broken areas in financial services”.

In addition, the company will expand its operations in the U.S., including building out executive and product teams in San Francisco.

“When we launched in the U.S it became quickly apparent that it would be our dominant market,” says the Cleo founder. “We were signing up about 1,000 users per day in the U.K. at the time. After [just] a week, we were at 10,000 per day in the U.S. and kept growing”.

Hussey-Yeo attributes a lot of the success state-side to better banking APIs in the U.S. with Plaid, which he says “dramatically drove up” conversion rates and lowered customer acquisition costs. “This combined with a much larger market focused us 100% on winning in the U.S. first. This race is far from played out”.

Hussey-Yeo says Cleo will always offer a free version “because everybody needs the ability to make smarter decisions about their financial lives”. In contrast, the premium product costs $5.99 per month and is intended for people who “need a little extra help,” providing features such as gamified savings, and “levelled up” credit scores with coaching. Premium users can also trigger a $100 salary advance designed to stop them dipping into a costly bank overdraft.

Meanwhile, Cleo says it has grown revenue by 400 percent in the last twelve months, and Hussey-Yeo tells me the company is now doing $10 million-plus AAR. In a further nod towards stronger unit economics, the cost to acquire a user is now less than $2 with the vast majority of users acquired organically.

“So we’re definitely not in growth at all costs mode,” he says. “We’ve worked hard over the last 12 months to bring our payback period comfortably under 12 months which is an incredibly rare feat at scale for fintech”.

One interesting aspect of Cleo’s mission and the strong “personality” its chatbot exhibits, is that the problem space the company is tackling is potentially based on behavioural science as much as it is data science. Hussey-Yeo doesn’t disagree.

“This is where the idea of a conversational interface came from,” he says. “I realised you could break the complexity down into a language anyone could engage with, and I mean actually look forward to engaging with it, to ultimately change their behaviour.

“Today we have an outstanding machine learning department but just as importantly is the behavioural researchers and writers at Cleo. We know this is the combination that makes Cleo special and, I hope, will eventually lead to us being the financial advisor for a billion people”.

German fintech Solactive, a technology-enabled provider of indices and index solutions to the global financial services industry, has taken a $60.4m / €50 million minority investment from growth investor Summit Partners.

Widely regarded as something of a disruptor in the global indexing industry, Solactive is benefitting from the ongoing and significant shift of active to passive investment strategies. Although Europe-based, it has more than 400 clients worldwide, including major investment banks, ETF providers and hedge funds, with more than $200 billion of global assets linked to more than 17,000 Solactive indices.

The idea behind Solactive is to take a customer-centric and technology-first approach to the creation, development, calculation and distribution of indices and related services. Because it is technology-first, it can offer tailor-made index solutions for ETFs and other index-linked investments across equity, fixed income and multi-asset class strategies.

Steffen Scheuble, the CEO of Solactive who founded Solactive in 2007, said: “We are proud to be one of the fastest-growing index providers in the world and widely regarded as a disruptor in the global index market. We are excited to join forces with Summit Partners and work together to further accelerate our journey.”

Summit brings significant experience in the technology and financial services sectors, as well as deep resources in supporting the expansion of our businesses across products, regions and client segments.”
Mr.

Johannes Grefe, a Principal with Summit Partners who will be joining the Solactive Board of Directors, commented: “Steffen and the Solactive team have recognized and responded to this trend with a highly automated and customer-centric approach based on a robust technology platform. With a strong combination of vision, product and client reach, we believe Solactive is well-positioned to continue its innovative disruption of the indexing market.”

Solactive serves its global customer base from its headquarters in Frankfurt am Main, Germany, and offices in Toronto, Hong Kong, Berlin and Dresden.

UK-based Orbex has raised a new $24 million funding round, led by BGF and Octopus Ventures, and including participation from existing investors High-Tech Gründerfonds, Heartcore Capital, and Elecnor S.A. This new investment help “secure the roadmap” that Orbex was already working towards regarding its launch vehicle development and deployment, which is currently targeting 2022 for its first commercial launches.

Orbex is set to launch its Orbex Prime rocket from the new Space Hub Sutherland spaceport in Scotland, and has six launch contracts on the books already. Its debut launch vehicle is a small payload rocket, which is unique in the industry in that it makes use of bio-propane, a fuel source that offers 90% emissions reduction vs. the traditional kerosene-based rocket fuels generally used for liquid rockets, and which is a renewable fuel source. Orbex also aims to reduce its impact in other ways, including through recoverability and reusability, and says that Prime’s design is intended to leave no post-launch debris either in the ocean on Earth, or even in orbit.

The startup is already in the process of building out its first Prime rockets, at two factories it operates in Scotland and in Denmark. The company uses 3D-printing heavily in its rocket facbbrication process. It has raised around $63 million to date, with its last round of around $39 million announced in 2018.

Productivity software has been getting a major re-examination this year, and human resources platforms — used for hiring, firing, paying, and managing employees — have been no exception. Today, one of the startups that’s built what it believes is the next generation of how HR should and will work is announcing a big fundraise, underscoring its own growth and the focus on the category.

Hibob, the startup behind the HR platform that goes by the name of “bob” (the company name is pronounced, “Hi, Bob!”), has picked up $70 million funding at a valuation that sources close to the company tell us is around $500 million.

“Our mission is to modernize HR technology,” said Ronni Zehavi, Hibob’s CEO, who cofounded the company with Israel David. “We are a people management platform for how people work today. Whether that’s remotely or physically collaborative, our customers face challenges with work. We believe that the HR platforms of the future will not be clunky systems, annoying, giant platforms. We believe it should be different. We are a system of engagement rather than record.”

The Series B is being led by SEEK and Israel Growth Partners, with participation also from Bessemer Venture Partners, Battery Ventures, Eight Roads Ventures, Arbor Ventures, Presidio Ventures, Entree Capital, Cerca Partners, and Perpetual Partners, the same group that also backed Hibob in its last round (a Series A extension) in 2019. It has raised $124 million to date.

The company has its roots in Israel but these days describes its headquarters as London and New York, and the funding comes on the back of strong growth. In an interview, Zehavi said that Hibob specialises in the mid-market customers and says that it has over 1,000 of them currently on its books across the US, Europe and Asia, including Monzo, Revolut, Happy Socks, Ironsource, Receipt Bank, Fiverr, Gong, and VaynerMedia. In the last year Hibob has had “triple digit” year-on-year growth (it didn’t specify what those digits are).

Human resources has never been at the more glamorous end of how a company works, and it can sometimes even be looked on with some disdain. However HR has found itself in a new spotlight in 2020, the year when every company — whether one based around people sitting at desks or in more interactive and active environments — had to change how it worked.

That might have involved sending everyone home to sign in from offices possibly made out of corners of bedrooms or kitchens, or that might have involved a vastly different set of practices in terms of when and where workers showed up and how they interacted with people once they did. But regardless of the implementations, they all involved a team of people who needed to be linked together and still feeling connected and managed; and sometimes hired, furloughed, or let go.

That focus has started to reveal the strains of how some legacy systems worked, with older systems built to consider little more than creating an employee identity number that could then be tracked for payroll and other purposes.

Hibob — Zehavi said they chose the name after the person who owned the bob.com domain wanted too much to sell it, but they liked bob for the actual product — takes an approach from the ground up that is in line with how many people work today, balancing different software and apps depending on what they are doing, and linking them up by way of integrations: its own includes Slack, Microsoft Teams, and Mercer and other packages that are popular with HR departments. 

While it covers all of the necessary HR bases like payroll and further compensation, onboarding, managing time off, and benefits, it further brings in a variety of other features that help build out bigger profiles of users, such as performance and culture, with the ability for peers, managers and workers themselves to provide feedback to enhance their own engagement with the company, and for the company to have a better idea of how they are fitting into the organization, and what might need more attention in the future.

That then links into a bigger organizational chart and conceptual charts that highlight strong performers, those who are possible flight risks, those who are leaders and so on. While there have been a number of others in the HR world that have built standalone apps that cover some of these features, what’s notable here is how they are all folded into one system together.

[gallery ids="2085416,2085417,2085418,2085419"]

The end effect, as you can see here, looks less like word salad and more interactive, graphic interfaces that are presumably a lot more enjoyable and at least easier to use for HR people themselves.

The importance for investors has been that the product and the startup has identified the opportunity, but has delivered not just more engagement, but a strong piece of software that still provides the essentials.

“This is certainly not a Workday,” said Adam Fisher, a partner at Bessemer, in an interview. “Our overall thesis has been that HR is only growing in importance. And while engagement is super important, that opportunity is not enough to create the market.”

The end result is a platform that has a significant shot at building in even more over time. For example, another large area that has been seeing traction in the world of enterprise and B2B software is employee training. Specifically, enterprise learning systems are creating another way to help keep people not only speed on important aspects of how the work, but also engaged at a time when connections are under strain.

“Training, a SuccessFactors-style offering, is definitely in our roadmap,” said Zehavi, who noted they are adding new features all the time. The latest has been compensation, sometimes known as merit increase cycles. “That is a very complex issue and requires deeper integrations finance and and the CFO’s office. We streamlined it and made it easy to use. We launched two months ago and it’s on fire. After learning and development the are other modules also down the road.”

Two months ago, Kubernetes observability platform Pixie Labs launched into general availability and announced a $9.15 million Series A funding round led by Benchmark, with participation from GV. Today, the company is announcing its acquisition by New Relic, the publicly traded monitoring and observability platform.

The Pixie Labs brand and product will remain in place and allow New Relic to extend its platform to the edge. From the outset, the Pixie Labs team designed the service to focus on providing observability for cloud-native workloads running on Kubernetes clusters. And while most similar tools focus on operators and IT teams, Pixie set out to build a tool that developers would want to use. Using eBPF, a relatively new way to extend the Linux kernel, the Pixie platform can collect data right at the source and without the need for an agent.

At the core of the Pixie developer experience are what the company calls “Pixie scripts.” These allow developers to write their debugging workflows, though the company also provides its own set of these and anybody in the community can contribute and share them as well. The idea here is to capture a lot of the informal knowledge around how to best debug a given service.

“We’re super excited to bring these companies together because we share a mission to make observability ubiquitous through simplicity,” Bill Staples, New Relic’s Chief Product Officer, told me. “[…] According to IDC, there are 28 million developers in the world. And yet only a fraction of them really practice observability today. We believe it should be easier for every developer to take a data-driven approach to building software and Kubernetes is really the heart of where developers are going to build software.”

It’s worth noting that New Relic already had a solution for monitoring Kubernetes clusters. Pixie, however, will allow it to go significantly deeper into this space. “Pixie goes much, much further in terms of offering on-the-edge, live debugging use cases, the ability to run those Pixie scripts. So it’s an extension on top of the cloud-based monitoring solution we offer today,” Staples said.

The plan is to build integrations into New Relic into Pixie’s platform and to integrate Pixie use cases with New Relic One as well.

Currently, about 300 teams use the Pixie platform. These range from small startups to large enterprises and as Staples and Asgar noted, there was already a substantial overlap between the two customer bases.

As for why he decided to sell, Pixie co-founder (and former Google AI CEO Zain Asgar told me that it was all about accelerating Pixie’s vision.

“We started Pixie to create this magical developer experience that really allows us to redefine how application developers monitor, secure and manage their applications,” Asgar said. “One of the cool things is when we actually met the team at New Relic and we got together with Bill and [New Relic founder and CEO] Lou [Cirne], we realized that there was almost a complete alignment around this vision […], and by joining forces with New Relic, we can actually accelerate this entire process.”

New Relic has recently done a lot of work on open-sourcing various parts of its platform, including its agents, data exporters and some of its tooling. Pixie, too, will now open-source its core tools. Open-sourcing the service was always on the company’s roadmap, but the acquisition now allows it to push this timeline forward.

“We’ll be taking Pixie and making it available to the community through open source, as well as continuing to build out the commercial enterprise-grade offering for it that extends the New Relic one platform,” Staples explained. Asgar added that it’ll take the company a little while to release the code, though.

“The same fundamental quality that got us so excited about Lew as an EIR in 2007, got us excited about Zain and Ishan in 2017 — absolutely brilliant engineers, who know how to build products developers love,” Bessemer Ventures General Partner Eric Vishria told me. “New Relic has always captured developer delight. For all its power, Kubernetes completely upends the monitoring paradigm we’ve lived with for decades.  Pixie brings the same — easy to use, quick time to value, no-nonsense approach to the Kubernetes world as New Relic brought to APM.  It is a match made in heaven.”

This morning Candu, a software startup that provides no-code web tools for SaaS apps, announced a $5 million funding round.

TechCrunch caught up with founder Jonathan Anderson and lead investor Villi Iltchev, a partner at Two Sigma Ventures, to chat about the deal.

First thing: This round, a bit like Scotch, has aged. It’s from March. And while it is very common for venture capital rounds to get announced after the fact, a nearly nine-month delay would normally be a stretch. However, Anderson explained that he wanted to hold off on releasing the news until Candu’s product was in the market. That’s a valid reason, so we’re going to chat about the investment despite its age.

The round we’re discussing today is not Candu’s first, as the startup previously raised around $1 million after its graduation from Entrepreneur First, an accelerator with a global footprint.

The $5 million investment began like many venture investments do, with an investor getting in touch with a young company months before a transaction is put together. Iltchev reached out to Candu in September of 2019 after hearing about the startup and liking its effort to allow for far-greater personalization of onboarding experiences inside of SaaS apps.

Anyone who has fired up a new business service, only to be given a walkthrough of how a few of the buttons work — a set of directions that I am sure you also skip — knows that the help can be generic, and annoying. Candu wants to work on the problem, offering no-code development tools to help non-engineers at companies build tailored pages to help with onboarding. What does that mean? With Candu, a SaaS app could offer different user cohorts nigh-personalized onboarding experiences.

So what? For SaaS companies big and small, effective onboarding is an important method of driving user engagement, a key metric for any modern software company. Why? Active users are less likely to churn, and more likely to expand spend on a service over time. So getting users activated well is no small task.

WalkMe and Pendo have built material business by approaching the issue from a different angle. WalkMe last raised $90 million last December, while Pendo picked up $100 million last October.

Candu’s project was interesting, and it had an obvious market, but the deal didn’t come together for a few months after the investor and startup met each other. Iltchev asked Candu to stay in touch, and toward the end of 2019 Anderson reached out to alert the investor that some of his peers were circling. That call led to the round that was announced today.

Superpowers

No-code startups have the potential to empower workers at companies who have to fight for engineering resources. If more groups in a company can build and deploy product, they will be able to move more quickly, and get more done. And they won’t have to beg the engineering team for help.

For that reason no-code startups — and other no-code and low-code projects — are incredibly interesting. In the case of Candu, if its service finds adoption, marketing and customer success teams may be able to create and tweak onboarding experiences without needing help from other teams. That would not only save time, but could lead to better results as more iterations would be possible.

Candu is not done figuring itself out, however. Co-founder Anderson told TechCrunch that his company is still working on solidifying its product-market fit. That is perfectly reasonable for a startup that is still grinding its way through early seed checks.

Asked what the company expects its annual contract value (ACV) to average out to, Anderson said that the market would guide the company, which only launched its service in late October. For now, Candu lists its prices starting from around $200 per month on up. Those targets provide some context about who it expects to sell to.

Regardless, it’s now up to the startup to scale its customer base, and turn those early customers into both revenue and learnings, which it can furrow back into its operations and product. Let’s see how far it can get with its March-era seed investment.

The worldwide shift to virtual workplaces has been a blessing and a curse to people with hearing impairments. Having office chatter occur in text rather than speech is more accessible, but virtual meetings are no easier to follow than in-person ones — which is why real-time captioning startup Ava has seen a huge increase in users. Riding the wave, the company just announced two new products and a $4.5 million seed round.

Ava previously made its name in the deaf community as a useful live transcription tool for real-life conversations. Start the app up and it would instantly hear and transcribe speech around you, color-coded to each speaker (and named if they activate a QR code). Extremely useful, of course, but when meetings stopped being in rooms and started being in Zooms, things got a bit more difficult.

“Use cases have shifted dramatically, and people are discovering the fact that most of these tools are not accessible,” co-founder and CEO Thibault Duchemin told TechCrunch.

And while some tools may have limited captioning built in (for example Skype and Google Meet), it may or may not be saved, editable, accurate, or convenient to review. For instance Meet’s ephemeral captions, while useful, only last a moment before disappearing, and are not specific to the speaker, making them of limited use for a deaf or hard of hearing person trying to follow a multi-person call. And the languages they are available in are limited as well.

As Duchemin explained, it began to seem much more practical to have a separate transcription layer that is not specific to any one service.

Thus Ava’s new product, a desktop and web app called Closed Captioning, which works with all major meeting services and online content, captioning it with the same on-screen display and making the content accessible via the same account. That includes things like YouTube videos without subtitles, live web broadcasts, and even audio-only content like podcasts, in more than 15 languages.

Individual speakers are labeled, automatically if an app supports it, like Zoom, or by having people in the meeting click a link that attaches their identity to the sound of their voice. (There are questions of privacy and confidentiality here, but they will differ case by case and are secondary to the fundamental capability of a person to participate.)

The transcripts all go to the person’s Ava app, letting them check through at their leisure or share with the rest of the meeting. That in itself is a hard service to find, Duchemin pointed out.

“It’s actually really complicated,” he said. “Today if you have a meeting with four people, Ava is the only technology where you can have accurate labeling of who said what, and that’s extremely valuable when you think about enterprise.” Otherwise, he said, unless someone is taking detailed notes — unlikely, expensive, and time-consuming — meetings tend to end up black boxes.

For such high quality transcription, speech-to-text AI isn’t good enough, he admitted. It’s enough to follow a conversation, but “we’re talking about professionals and students who are deaf or hard of hearing,” Duchemin said. “They need solutions for meetings and classes and in-person, and they aren’t ready to go full AI. They need someone to clean up the transcript, so we provide that service.”

Ava Scribe quickly brings in a human trained not in direct transcription but in the correction of the product of speech-to-text algorithms. That way a deaf person attending a meeting or class can follow along live, but also be confident that when they check the transcript an hour later it will be exact, not approximate.

Right now transcription tools are being used as value-adds to existing products and suites, he said — ways to attract or retain customers. They aren’t beginning with the community of deaf and hard of hearing professionals and designing around their needs, which is what Ava has striven to do.

The explosion in popularity and obvious utility of their platform has led to this $4.5M seed round, as well, led by Initialized Capital and Khosla Ventures.

Duchemin said they expected to double the size of their team with the money, and start really marketing and finding big customers. “We’re very specialized, so we need a strong business model to grow,” he said. A strong, unique product is a good place to start, though.

Papercup, the U.K.-based AI startup that has developed speech technology that translates people’s voices into other languages and is already being used in the video and television industry, has raised £8 million in funding.

The round was led by LocalGlobe and Sands Capital Ventures, alongside Sky, GMG Ventures, Entrepreneur First (EF) and BDMI. Papercup says the new capital will be used to invest further into machine learning research and to expand its “human-in-the-loop” quality control functionality, which is used to improve and customise the quality of its AI-translated videos.

Meanwhile, Papercup’s existing angel investors include William Tunstall-Pedoe, the founder of Evi Technologies — the company acquired by Amazon to create Alexa — and Zoubin Ghahramani, former chief scientist and VP of AI at Uber and now part of the Google Brain leadership team.

Founded in 2017 by Jesse Shemen and Jiameng Gao while going through EF’s company builder program, Papercup is building out an AI and machine learning-based system that it says is capable of translating a person’s voice and expressiveness into other languages. Unlike a lot of text-to-speech, the startup claims the resulting voice translation is “indistinguishable” from human speech, and, perhaps uniquely, it attempts to retain the characteristics of the original speaker’s voice.

Initially, the tech is being targeted at video producers, including already being used by Sky News, Discovery and YouTube stars Yoga with Adriene, along with DIY content creators. It is pitched as a much more scalable and therefore lower-cost alternative to pure human dubbing.

“Most of the world’s video and audio content is shackled to a single language,” says Papercup co-founder and CEO Shemen. “That includes billions of hours of videos on YouTube, millions of podcast episodes, tens of thousands of classes on Skillshare and Coursera, and thousands of hours of content on Netflix. Almost every content owner is scrambling to go international, but there is yet no simple and cost-effective way to translate content beyond subtitling”.

For “deep pocketed studios,” there is of course the option to employ high-end dubbing via a professional dubbing studio and voice actors, but this is far too expensive for most content owners. And even wealthy studios are often constrained in terms of how many languages they can accommodate.

“That leaves the mid and long tail of content owners — literally 99% of all content — stranded and incapable of reaching international audiences beyond subtitling,” says Shemen, which, of course, is where Papercup comes into play. “Our aim is to generate translated voices that sound as close to the original speaker as possible”.

To do that, he says that Papercup will need to tackle four things. First up is creating “natural sounding” voices, i.e. how clear and human-like the synthetic voices sound. The second challenge is retaining emotion and pacing to reflect how the original speaker expressed themselves (think: happy, sad, angry etc.). Third is capturing the uniqueness of someone’s voice (e.g. Morgan Freeman, but in German). Lastly, the resulting translation needs the correct alignment of the audio to the video itself.

Explains Shemen: “We started off by making our voices as human-like and natural sounding as possible, where we’ve made quite a significant leap in terms of quality by honing our technology to the task, and today we have one of the best Spanish speech synthesis systems in production.

“We’re now focusing on better retainment and transfer of the original emotion and expressiveness in the original speaker across languages, and meanwhile figuring out what it is exactly that makes for quality dubbing”.

The next challenge and arguably the toughest nut to crack is “speaker adaptation,” described as capturing the uniqueness of someone’s voice. “This is the last layer of adaptation,” notes the Papercup CEO, “but it was also one of our first breakthroughs in our research. While we have models that can accomplish this, we’re focusing more of our time on emotion and expressiveness”.

That’s not to say Papercup is entirely machine-powered, even if it might be one day. The company also employs a “human-in-the-loop” process to make corrections and adjustments to the translated audio track. This includes correcting for any speech recognition or machine translation errors that come up, making adjustments to the timings of the audio, as well as enforcing emotions (e.g. happy, sad) and changing the speed of the generated voice.

How much human-in-the-loop is required depends on the type of content and priorities of the content owners, i.e. how realistic or perfect they need the resulting video to be. In other words, it isn’t a zero-sum game, as good enough will be more than enough for a swathe of content owners at scale.

Asked about the technology’s beginnings, Shemen says Papercup started with research conducted by co-founder and CTO Jiameng Gao “who is incredibly smart and oddly obsessed with speech processing”. Gao completed two Masters at University of Cambridge (in machine learning and speech language technology) and wrote a thesis on speaker adaptive speech processing. It was at Cambridge that he realised that something like Papercup was possible.

“When we started working together at Entrepreneur First at the end of 2017, we built our initial prototype systems that showed that this technology was even possible despite there being no precedent for it,” says Shemen. “Based on early conversations, the demand was clearly overwhelming for what we were building — it was just a function of actually building something that could be used in a production environment”.

Another batch of complaints has been filed with European Union data protection agencies urging enforcement action against the adtech industry’s abuse of Internet users’ information to target ads.

The complaints argue that behavioural ads are both harmful and unlawful.

Earlier complaints over the same Real-Time Bidding (RTB) programmatic advertising issue were filed across the EU in 2018 and 2019 but have yet to result in any substantive regulatory action.

Ireland did open a probe into Google’s ad exchange last year. While Belgium’s DPA has been progressing an investigation into a flagship industry tool that’s used for gathering consents to ad targeting — making a preliminary finding of non-compliance in October. But litigation to reach a final verdict on the IAB Europe’s ‘Transparency and Consent’ (TCF) framework won’t take place until next year.

(Related: The UK’s data protection agency is facing a legal challenge over its failure to act on RTB complaints, despite repeatedly expressing concern about the industry’s lawfulness problem.)

Both Google and the IAB continue to deny any problems with their adtech. Last year Google said authorised buyers that use its systems are subject to “stringent policies and standards”. While the IAB Europe rejected the Belgium DPA’s findings — saying its preliminary report “fundamental misunderstand[s]” the TCF tech.

The latest GDPR complaints target how the RTB component of programmatic advertising broadcasts Internet users’ personal data to scores of entities involved in these high speed eyeball auctions — arguing it runs counter to core security requirements in the General Data Protection Regulation (GDPR), as well as being horrible for people’s privacy.

A key principle of the GDPR is security by design and default — with the regulation placing legal requirements on personal data handlers to make sure people’s information is properly secured.

The complaints, which target Google and the IAB in their capacity as RTB standard setters, have been filed by civil society groups in six European countries — namely: Asociatia pentru Tehnologie si Internet (ApTi), Romania; D3 – Defesa dos Direitos Digitais, Portugal; GONG, Croatia; Global Human Dignity Foundation, Malta; Homo Digitalis, Greece; and the Institute of Information Cyprus.

They’re being coordinated by a consortium led by the Civil Liberties Union for Europe (Liberties), the ORG (Open Rights Group) and the Panoptykon Foundation. 

“Real-time bidding, which is the bedrock of the online advertising industry, is an abuse of people’s right to privacy,” said Dr Orsolya Reich, senior advocacy officer at Liberties, in a supporting statement. “The GDPR has been in place since 2018 and it is there precisely to give people a greater say about what happens to their data online.

“Today, more civil society groups are saying enough with this invasive advertising model and are asking data protection authorities to stand up against the harmful and unlawful practices they use.”

The consortium is asking for a joint investigation by their respective national DPAs — and for regulators to join with ongoing adtech investigations in Ireland (into Google’s adtech) and Belgium (into the IAB Europe’s TCF framework).

It’s not clear how far the Irish DPC’s investigation of Google has progressed — but it continues to face criticism for the lack of decisions on cross-border GDPR cases, some 2.5 years after the regulation technically begun being applied.

A mechanism in the GDPR means cross-border cases (basically anything related to mainstream consumer tech) get passed to a lead agency to investigate. However other agencies also remain involved, as interested parties, and must agree with any final decision made.

The system has led to a bottleneck of cases in certain EU locations, such as Ireland, where many tech giants base their European HQ. So the concern is this one-stop-shop mechanism is adding an unworkable level of friction to GDPR investigations — delaying decisions and enforcement action so much it risks the entire framework.

The Commission has acknowledged weakness in GDPR enforcement. Most obviously because it’s working on a massive package of new digital regulations. Though its strategy for fixing the enforcement problem is less clear as EU Member States look set to remain responsible for the bulk of this additional oversight, just as they’re responsible for resourcing their own DPAs now. (And yet more complaints have been filed this year accusing European governments of a GDPR resourcing failure.)

Ireland’s DPC is slated to issue its first cross-border GDPR decision in a case that relates to a Twitter security breach very shortly. But last year its commissioner, Helen Dixon, suggested it would come with its first such decisions early in 2020 — so the gap between GDPR expectation and reality is running almost 12 months late at this point.

 

The consortium filing the latest RTB complaints writes in a press release that while some of the earlier adtech complaints were referred to lead authorities it has no knowledge of “any meaningful cooperation or joint operations between national authorities and the lead authorities”.

“This suggests that cooperation and consistency mechanisms as envisioned in the GDPR are yet to be implemented fully,” the group adds, calling for a joint investigation into the RTB issue because the technology functions in the same way across borders — and “produces the same negative effects in all EU member states”, as they put it.

However it’s not clear how extra joint working — if indeed that’s really what’s being called for — would help to speed up GDPR enforcement. Nor how referring additional complaints to Ireland and Belgium would work to speed up their current investigations.

Most likely, the intent is to keep up pressure on the regulators to act.

Asked about the call for joint working, a Liberties spokesman told us: “The problem is that Google and IAB are big players, standard-setters in the market, and they affect all Internet users. Given the geographical scope of the issues raised in the complaints, we think it’s better for supervisory authorities to act in unison, not to be working alone in their corner.  This is why national partners are inviting their national DPAs to refer this complaint to the lead supervisory authorities who are already investigating Google’s and IAB’s compliance with the GDPR.”

Commenting in another supporting statement, Mariano delli Santi, legal and policy officer at the ORG, added: “These new complaints show that the GDPR is working. Individuals are increasingly aware of their rights, and they demand change. Now, it is up to the authorities to support this process, and make sure these laws are properly and consistently enforced against the widespread abuses of the adtech industry.”

At the time of writing, the only extant example of enforcement against a tech giant under the updated regulation was a January 2019 decision to fine Google $57M by France’s CNIL. That investigation was limited to having a national scope, though, rather than being treated as a cross-border case.

Since then Google has shifted its legal base in Europe to Ireland — so now falls under the lead jurisdiction of the DPC.

This arrangement appears to suit big tech, enabling it to avoid the risk of speedier investigations conducted by single Member State agencies acting faster alone. (So it’s very interesting to see TikTok ramping up its business infrastructure and headcount in Ireland — as it’s also now on CNIL’s radar… )

 

As noted earlier, EU lawmakers have conceded GDPR enforcement has been a weakness thus far.

In a review of the two year old regulation this summer the Commission highlighted a lack of universally vigorous enforcement.

Last week the values and transparency commissioner, Vera Jourouv, also raised the problem as she set out the bloc’s plan to bolster democratic values against a range of online risks, such as algorithmically amplified or microtargeted disinformation and election interference — acknowledging GDPR alone isn’t enough to fix myriad intersecting tech-fuelled problems.

“[After the Cambridge Analytica scandal] we said that we are relieved that after GDPR came into force we are protected against this kind of practice — that people have to give consent and be aware of that — but we see that it might be a weak measure only to rely on consent or leave it for the citizens to give consent,” she said.

“Enforcement of privacy rules is not sufficient — that’s why we are coming in the European Democracy Action Plan with the vision for the next year to come with the rules for political advertising, where we are seriously considering to limit the microtargeting as a method which is used for the promotion of political powers, political parties or political individuals.”

The European Commission is in the progress of drafting an ambitious and interlocking package of digital regulations, that it wants to fuel a regional data economy and set firm online rules to engender the necessary trust — and has said it wants this major digital policymaking effort to serve Europe for decades.

But without effective enforcement of its Internet rulebook it’s not clear how the bloc’s digital strategy will deliver as intended.