Month: June 2019

The old guard of advertising networks continues to swap hands. In the latest development, marketing company Fluent has announced that it is buying AdParlor, a company that built early inroads into advertising and advertising strategy on social media networks like Facebook and Twitter. According to an 8-K filing, the deal is valued at $10 million: $7.5 million in cash when the acquisition closes, and a further $2.5 million over the next two years.

Fluent focuses on performance marketing solutions based around what it says are a network of its own, proprietary websites that collectively get about 1 million visitors daily, along with data for some 190 million opted-in consumers who fill out surveys and are then part of a bigger marketing database that can be used for targeting on behalf of its customers.

The plan is to integrate AdParlor’s technology and services to its larger stack of mar-tech tools. “I am thrilled to welcome Evan Conway and the AdParlor team,” said Ryan Schulke, CEO of Fluent, said in a statement. “The combination of their high-touch approach to managing digital strategies with tech-enabled media buying and our scalable, first-party, self-declared data asset creates a unique value proposition for performance-focused and growth-minded brands.”

The landing at Fluent is the latest chapter for a company that was once estimated to be Facebook’s biggest Ads API vendor — albeit at a time when advertising on the social network was just starting to take off, with key clients of that moment including Groupon (it managed all of its Facebook advertising at a time when Groupon was a huge business, no small thing).

That traction made it an attractive acquisition for AdKnowledge, which acquired the company in 2011 amid a much larger shopping spree that saw it pick up a number of other small ad tech companies. The bigger strategy at the time for AdKnowledge was to build out its ad network, with a focus on offering inventory across the long tail of the internet beyond search and the very biggest websites. The price of that deal was never disclosed. AdParlor had always been bootstrapped prior to getting acquired but had grown to handle billions of impressions.

Online advertising has evolved, however, and companies like Google and Facebook (and from the looks of it, soon Amazon) have moved also to target the long tail, and it seems that AdKnowledge’s own business shifted focus, too. The company quietly rebranded some time ago to V2 Ventures and appears to be pursuing a longer-term strategy of divesting many of the businesses that it had acquired.

That brings us to Fluent. Coincidentally, similar to how AdKnowledge acquired AdParlor to expand into social, so too will Fluent be using the asset for a similar end, it seems: the plan will be to fold social media strategy, planning and buying into its wider services.

“We have worked hard to build a team and a company that specializes in delivering performance for our customers on social networks, and we are excited to be joining the Fluent family,” said Evan Conway, CEO of AdParlor, in a statement. “Our customer-focused approach will be enhanced by leveraging Fluent’s proprietary technology and resources, allowing us to provide more value across social channels, and leverage greater personalization and measurement capabilities.”

AdParlor changing hands is not the only M&A move we’ve seen of a legacy ad tech company in recent times. Amazon is currently in the process of acquiring bankrupt Sizmek’s Ad Server business for around $30 million (beating out at least one other party that was also interested, Maurice and Alain Levy’s Weborama).

The bigger picture here is that there are still definitely new opportunities to be identified and seized in the world of ad tech, but likely winners (or at least consolidators) will be those who already have larger strategies and perhaps even businesses in place, where smaller operations like AdParlor — or indeed Sizmek’s Ad Server business — will serve to fill obvious holes along the way.

Pretty much everything about making a self-driving car is difficult, but among the most difficult parts is making sure the vehicles know what pedestrians are doing — and what they’re about to do. Humanising Autonomy specializes this, and hopes to become a ubiquitous part of people-focused computer vision systems worldwide.

The company has raised a $5.3 million seed round from an international group of investors on the strength of its AI system, which it claims outperforms humans and works on images from practically any camera you might find in a car these days.

HA’s tech is a set of machine learning modules trained to identify different pedestrian behaviors — is this person about to step into the street? Are they paying attention? Have they made eye contact with the driver? Are they on the phone? Things like that.

The company credits the robustness of its models to two main things. First, the variety of its data sources.

“Since day one collected data from any type of source — CCTV cameras, dash cams of all resolutions, but also autonomous vehicle sensors,” said co-founder and CEO Maya Pindeus. “We’ve also built data partnerships and collaborated with different institutions, so we’ve been able to build a robust data set across different cities with different camera types, different resolutions and so on. That’s really benefited the system, so it works in nighttime, rainy Michigan situations, etc.”

Notably their models rely only on RGB data, forgoing any depth information that might come from lidar, another common sensor type. But Pindeus said that type of data isn’t by any means incompatible, it just isn’t as plentiful or relevant as real-world, visual-light footage.

In particular HA was careful to acquire and analyze footage of accidents, since these are especially informative cases of failure of AVs or human drivers to read pedestrian intentions, or vice versa.

The second advantage Pindeus claimed is the modular nature of the models the company has created. There isn’t one single “what is that pedestrian doing” model, but a set of them that can be individually selected and tuned according to the autonomous agent’s or hardware’s needs.

“For instance, if you want to know if someone is distracted as they’re crossing the street. There’s a lot of things that we do as humans to tell if someone is distracted,” she said. “We have all these different modules that kind of come together to predict whether someone’s distracted, at risk, etc. This allows us to tune it to different environments, for instance London and Tokyo – people behave differently in different environments.”

“The other thing is processing requirements; Autonomous vehicles have a very strong GPU requirement,” she continued. “But because we build in these modules, we can adapt it to different processing requirements. Our software will run on a standard GPU when we integrate with level 4 or 5 vehicles, but then we work with aftermarket, retrofitting applications that don’t have as much power available, but the models still work with that. So we can also work across levels of automation.”

The idea is that it makes little sense to aim only for the top levels of autonomy when really there are almost no such cars on the road, and mass deployment may not happen for years. In the meantime, however, there are plenty of opportunities in the sensing stack for a system that can simply tell the driver that there’s a danger behind the car, or activate automatic emergency braking a second earlier than existing systems.

While there are lots of papers published about detecting pedestrian behavior or predicting what a person in an image is going to do, there are few companies working specifically on that task. A full stack sensing company focusing on lidar and RGB cameras needs to complete dozens or hundreds of tasks, depending on how you define them: object characterizations and tracking, watching for signs, monitoring nearby and distant cars, and so on. It may be simpler for them and for manufacturers to license HA’s functioning and highly specific solution rather than build their own or rely on more generalized object tracking.

“There are also opportunities adjacent to autonomous vehicles,” pointed out Pineus. Warehouses and manufacturing facilities use robots and other autonomous machines that would work better if they knew what workers around them were doing. Here the modular nature of the HA system works in its favor again — retraining only the parts that need to be retrained is a smaller task than building a new system from scratch.

Currently the company is working with mobility providers in Europe, the U.S., and Japan, including Daimler Mercedes Benz and Airbus. It’s got a few case studies in the works to show how its system can help in a variety of situations, from warning vehicles and pedestrians about each other at popular pedestrian crossings to improving path planning by autonomous vehicles on the road. The system can also look over reams of past footage and produce risk assessments of an area or time of day given the number and behaviors of pedestrians there.

The $5M seed round, led by Anthemis, with Japan’s Global Brain, Germany’s Amplifer, and SV’s Synapse Partners, will mostly be dedicated to commercializing the product, Pineus said.

“The tech is ready, now it’s about getting it into as many stacks as possible, and strengthening those tier 1 relationships,” she said.

Obviously it’s a rich field to enter, but still quite a new one. The tech may be ready to deploy but the industry won’t stand still, so you can be sure that Humanising Autonomy will move with it.

Seventy-two hours to save. That’s how much time remains on super-early-bird pricing for passes to Disrupt San Francisco 2019, which takes place October 2-4. If you plan to attend TechCrunch’s flagship event dedicated to bold, early-stage startuppers — and why wouldn’t you — you have until June 21 at 11:59 p.m. (PT) to score the best price. Depending on the type of pass you buy, you can save serious cheddar — up to $1,800. Choose the budget-friendly payment plan option during checkout and you can pay for your pass over time. It’s all geared to be as easy on the purse strings as possible, so buy your pass now and save.

Moscone North Convention Center will be home to more than 10,000 attendees from around the world — including startups, exhibitors and media outlets — for three jam-packed days of programming across 14 categories. It’s where you’ll find both the present and future of technology under one roof.

Every Disrupt event features an amazing lineup of speakers. We’re talking interviews and panel discussions with some of the world’s most influential names in tech and investing — and that tradition continues at Disrupt SF 2019. Here’s just one example of the presentations you’ll experience.

Cybersecurity ranks as a major concern that affects everyone — consumers, businesses and governments. We’re thrilled that Jeanette Manfra, homeland security assistant director and a senior executive at the department’s Cybersecurity and Infrastructure Security Agency (CISA), will grace the stage. One of the government’s most experienced cybersecurity civil servants, Manfra currently leads the effort to protect and strengthen our nation’s vital infrastructure, including the power grid and water supplies. We can’t wait to hear what she has to say about the government’s cybersecurity efforts.

Experience Startup Alley’s ocean of opportunity. You’ll find hundreds of outstanding early-stage startups pushing tech boundaries and creating the future today. You’ll also find the TC Top Picks — a hand-picked cadre of companies representing these categories: AI/Machine Learning, Biotech/Healthtech, Blockchain, Fintech, Mobility, Privacy/Security, Retail/E-commerce, Robotics/IoT/Hardware, SaaS and Social Impact & Education. If that describes your early-stage startup, apply to be a TC Top Pick. If you make the cut, you get to exhibit in Startup Alley for free. You also get three complimentary passes and VIP treatment with plenty of VC and media exposure — including a live interview with a TechCrunch editor on the Showcase Stage.

Disrupt SF 2019 takes place October 2-4, and your chance to snag the best price on passes disappears in just three short days. Be bold. Buy your pass now — before the June 21 11:59 p.m. (PT) deadline — and save big.

Is your company interested in sponsoring or exhibiting at Disrupt SF 2019? Contact our sponsorship sales team by filling out this form.

Future autonomous vehicle and other machine intelligence systems might not need line-of-sight to gather incredibly detailed image data: New research from Carnegie Mellon University, the University of Toronto and University College London has devised a technique for “seeing around corners.”

The method uses special sources of light, combined with sensors and computer vision processing to effectively infer or rebuild extremely detailed imagery, much more detailed than has been possible previously, without having photographed it or otherwise ‘viewed’ it directly.

There are some limitations – so far, researchers working on the project have only been able to use this technique effectively for “relatively small areas,” according to CMU Robotics Institute Professor Srinivasa Narasimhan.

That limitation could be mitigated by employing this technique alongside others used in the field of non-line-of-site (or NLOS) computer vision research. Some such techniques are already in use in the market, including how Tesla’s Autopilot system (and other driver-assist technologies) makes use of reflected or bounced radar signals to see around the cars immediately in front of the Tesla vehicle.

The technique used in this new study is actually similar to what happens in a LiDAR system used in many autonomous vehicle systems (though Tesla famously eschews use of laser-based vision systems in its tech stack). CMU and its partner institutions use ultrafast laser light in their system, bouncing it off a wall to light an object hidden around a corner.

Sensors then capture the reflected light when it bounces back, and researchers measure and calculate how long it took for the reflected light to return to the point of origin. Taking a number of measurements, and using information regarding the target object’s geometry, the team was able to then reconstruct the objects with remarkable accuracy and detail. Their method was so effective that it even works through semi-osbscruingf materials, including heavy paper – another big benefit when it comes to its potential for use in environment sensors that work in real-world conditions.

At left, an image of a quarter scanned using non-line-of-sight imaging. At right, an image of a quarter scanned using line-of-sight imaging.

TextIQ, a machine learning system that parses and understands sensitive corporate data, has raised $12.6 million in Series A funding led by FirstMark Capital, with participation from Sierra Ventures.

TextIQ started as cofounder Apoorv Agarwal’s Columbia thesis project titled “Social Network Extraction From Text.” The algorithm he built was able to read a novel, like Jane Austen’s Emma, for example, and understand the social hierarchy and interactions between characters.

This people-centric approach to parsing unstructured data eventually became the kernel of TextIQ, which helps corporations find what they’re looking for in a sea of unstructured, and highly sensitive, data.

The platform started out as a tool used by corporate legal teams. Lawyers often have to manually look through troves of documents and conversations (text messages, emails, Slack, etc.) to find specific evidence or information. Even using search, these teams spend loads of time and resources looking through the search results, which usually aren’t as accurate as they should be.

“The status quo for this is to use search terms and hire hundreds of humans, if not thousands, to look for things that match their search terms,” said Agarwal. “It’s super expensive, and it can take months to go through millions of documents. And it’s still risky, because they could be missing sensitive information. Compared to the status quo, TextIQ is not only cheaper and faster but, most interestingly, it’s much more accurate.”

Following success with legal teams, TextIQ expanded into HR/compliance, giving companies the ability to retrieve sensitive information about internal compliance issues without a manual search. Because TextIQ understands who a person is relative to the rest of the organization, and learns that organization’s ‘language’, it can more thoroughly extract what’s relevant to the inquiry from all that unstructured data in Slack, email, etc.

More recently, in the wake of GDPR, TextIQ has expanded its product suite to work in the privacy realm. When a company is asked by a customer to get access to all their data, or to be forgotten, the process can take an enormous amount of resources. Even then, bits of data might fall through the cracks.

For example, if a customer emailed Customer Service years ago, that might not come up in the company’s manual search efforts to find all of that customer’s data. But since TextIQ understands this unstructured data with a person-centric approach, that email wouldn’t slip by its system, according to Agarwal.

Given the sensitivity of the data, TextIQ functions behind a corporation’s firewall, meaning that TextIQ simply provides the software to parse the data rather than taking on any liability for the data itself. In other words, the technology comes to the data, and not the other way around.

TextIQ operates on a tiered subscription model, and offers the product for a fraction of the value they provide in savings when clients switch over from a manual search. The company declined to share any further details on pricing.

Former Apple and Oracle General Counsel Dan Cooperman, former Verizon General Counsel Randal Milch, former Baxter International Global General Counsel Marla Persky, and former Nationwide Insurance Chief Legal and Governance Officer Patricia Hatler are on the advisory board for TextIQ.

The company has plans to go on a hiring spree following the new funding, looking to fill positions in R&D, engineering, product development, finance, and sales. Cofounder and COO Omar Haroun added that the company achieved profitability in its first quarter entering the market and has been profitable for eight consecutive quarters.

More Americans than ever before are suffering from mental and emotional distress. In the U.S., the mental health problem is exacerbated by issues across infrastructure, government, and culture. However, because the resources for those living with mental health issues are constrained, startups could have a big impact.

In particular, we believe that digital therapeutic-approaches offer great promise in overcoming the problems inherent in traditional approaches to mental and behavioral therapy. Such problems relate to stigma, cost, and general inaccessibility of cost-effective treatments for the general population.

We are starting to see new energy behind innovators in the mental health space. Examples include Enlyte (discussed in greater detail below); Talkspace, an online therapy app that connects users with licensed therapists; Calm, a sleep and meditation app; and Feel, a wearable designed to monitor the user’s emotional state. Other examples are listed in the Appendix at the back.

Each of these companies—whether they aim to provide easy access to mental health professionals, to promote overall mental wellness, or to better monitor the user’s mental state—has the potential to be highly impactful as well as profitable.

In our view, the time is right to invest in mental health and digital therapeutics. In this paper, we provide an overview of the field of digital therapeutics for mental health, as well as the legal, regulatory and ethical issues that should be considered by entrepreneurs and investors.

Table of Contents

• The big four mental health afflictions: Stress, anxiety, addiction, depression
  • Stress
  • Anxiety
  • Addiction
  • Depression
• The promise of digital therapeutics
• FDA regulatory environment: Software as a medical device
  • Legislative exclusions of software (21st Century Cures Act)
  • Mobile apps
• Intellectual property protection
• Example startup: Enlyte
• Reasons to invest in digital therapeutics
• Appendix: Example startups
  • Pear Therapeutics
  • Stop, Breathe & Think
  • Pzizz
  • Simple Habit
  • Aura Health
  • Pacifica
  • 2Morrow

The big four mental health afflictions: Stress, anxiety, addiction, depression

Image via Getty Images / Feodora Chiosea

The mental health crisis costs companies around the world over $1 Trillion in lost productivity and increased health care insurance premiums annually. The productivity losses are primarily caused by absenteeism, and turnover and replacement costs.

In addition, the costs attributable to the family members and loved ones of employees (employee ecosystem) cost employers approximately 250% more in lost productivity than their direct employees. According to a report from the World Health Organization (WHO), 450 million people are currently suffering from mental health disorders leading to illness and disability.

The Lancet Commission on mental health predicts that by 2030, mental illness will cost the world USD 16 trillion. If we look at the US alone, 40.3 million people are affected by the disease of addiction. Twenty percent of US deaths are attributed to addiction to tobacco, alcohol, drugs and other substances.

Stress

The National Institute for Health (NIH) defines stress as a “physiological response to challenge or demand”. There are two forms of stress: acute and chronic. Acute stress is frequently referred to as your body’s fight-or-flight response.

Since 2001, Apple has staked its claim across the world with its own first-party brick and mortar locations. But the U.S. is a big country, and the 270 or so stores can only cover so much ground. In the past three years, the company says it has expanded repair coverage to three times as many locations in this massive country of ours, courtesy of third-party partnerships.

That list now includes almost 1,000 Best Buys, which now offer Apple certified repairs courtesy of 7,600 “newly Apple-certified technicians” capable of offering up same day repairs on iPhones and other products.

“At Apple, we’re dedicated to providing the best customer service in the world,” Apple Care VP Tara Bunch said in a release tied to this morning’s news. “If a customer ever needs to repair their products, we want them to feel confident those repairs are done safely and correctly. We’re always looking at how we can reliably expand our network of trained technicians and we’re excited to partner with every Best Buy store so it’s even easier for our customers to find an authorized repair location near them.”

It’s a deal that makes sense for both parties. For Apple, it means covering customers in locations like Yuma, Sioux City and Bismarck. This brings its total third-party authorized service locations up to 1,800 in the U.S.

For Best Buy, the deal means a partnership and blessing from another key electronics giant, with Apple joining the likes of Samsung, which currently has authorized Galaxy repairs from the big box store. More info on Apple repair services here.

China-based Zhejiang Geely Holdings is tapping Zenuity, the joint venture between Volvo and Swedish technology company Veoneer, as its preferred driver assistance and autonomous vehicle software supplier for its range of car brands.

The decision means Zenuity’s software will end up in Geely Auto, Geometry, Volvo Cars, performance brand Polestar, British carmaker Lotus and the subscription-based automaker Lynk & Co. Geely Holdings’s total group sales last year were 2.15 million vehicles, according to the company.

The supplier partnership solidifies business for Zenuity as it goes up against more established players in the industry. And it also hints at where some of Geely’s car brands are headed.

Zenuity also confirmed Wednesday that it plans to provide software that will allow unsupervised driving for Volvo’s next generation cars. Zenuity did not provide a timeline.

Geely Holdings acquired Volvo Cars in 2010 from Ford. Two years ago, Volvo and Autoliv-spinout Veoneer formed a joint venture called Zenuity with a focus on developing self-driving vehicle software.

Zenuity plans to demonstrated Wednesday its ADAS technology in a moving vehicle for the first time to illustrate its commercial readiness. The demonstration, held in Detroit at Veoneer’s Tech Day, will put a vehicle equipped with camera-only automatic emergency braking through several accident scenarios. The system is designed to meet future European vehicle safety regulations coming in 2020 and 2022.

Earlier this year, Zenuity received approval to test self-driving Volvos on Swedish highways. The tests will be done by trained drivers, with their hands off the steering wheel at a maximum speed of 50 miles per hour.

China is losing its global lead in games. By the end of 2019, the U.S. will replace China as the world’s largest gaming market with an estimated revenue of $36.9 billion, says a new report from research firm Newzoo.

This will mark the first time since 2015 that the U.S. will top the global gaming market, thanks to healthy domestic growth in consoles. Globally, Xbox, PlayStation, Nintendo and other console games are on track to rise 13.4% in revenue this year. Driving the growth is the continued shift toward the games-as-a-service model, Newzoo points out, on top of a solid installed base across the current console generation and spending from new model releases.

China, on the other hand, suffered from a nine-month freeze on game licenses last year that significantly shrank the stream of new titles. Though applications have resumed, industry experts warn of a slower and stricter approval process that will continue to put the squeeze on new titles. Time limits imposed on underage players will also hurt earnings in the sector.

As a result of China’s slowdown, Asia-Pacific is no longer the fastest-growing region. Taking the crown is Latin America, which is enjoying a 10.4% compound annual growth.

Despite China’s licensing blackout, Tencent remained as the largest publicly-listed gaming firm in 2018, pocketing $19.73 billion in revenue. Growth slowed to 9% compared to 51% from 2016 to 2017 at Tencent’s gaming division, but the Shenzhen-based company is back on track with new blockbuster Game for Peace (和平精英), a regulator-friendly version of PlayerUnknown’s Battleground, ready to monetize.

Trailing behind Tencent in the global ranking is Sony, Microsoft, Apple and Activision Blizzard.

Other key trends of the year:

Rise of instant games: Mini games played inside WeChat without installing another app are becoming mainstream in China. These games, which tend to have strong social elements and easy to play, have attracted followers including Douyin (TikTok’s Chinese version) to create with their own offerings.

Facebook’s Instant Games have also come a long way since opening to outside developers in 2018. The platform now sees more than 30 billion game sessions played across over 7,000 titles. WeChat doesn’t use the same metrics but for some context, the Chinese company boasted 400 million monthly players on mini games as of January.

Mobile momentum carries on: Mobile games will continue to outpace growth on PC and console in the coming years. As expected, emerging markets that are mobile-first and mobile-only will drive most of the boom in mobile gaming, which is on course to account for almost half (49%) of the entire sector by 2022. Part of the growth is driven by improved hardware and internet infrastructure, as well as a growing number of cross-platform titles.

Games in the cloud are here: It was a distant dream just a few years ago — being able to play some of the most demanding titles regardless of what hardware one owns. But the technology is closer than ever to coming true with faster internet speed and the imminent rollout of 5G networks. A few giants have already showcased their cloud gaming services over the last few months, with the likes of Google’s Stadia, Microsoft’s xCloud, and Tencent’s Start slated to test the market.