All posts in “articles”

Peltarion raises $20M for its AI platform

Peltarion, a Swedish startup founded by former execs from companies like Spotify, Skype, King, TrueCaller and Google, today announced that it has raised a $20 million Series A funding round led by Euclidean Capital, the family office for hedge fund billionaire James Simons. Previous investors FAM and EQT Ventures also participated, and this round brings the company’s total funding to $35 million.

There is obviously no dearth of AI platforms these days. Peltarion focus on what it calls “operational AI.” The service offers an end-to-end platform that lets you do everything from pre-processing your data to building models and putting them into production. All of this runs in the cloud and developers get access to a graphical user interface for building and testing their models. All of this, the company stresses, ensures that Peltarion’s users don’t have to deal with any of the low-level hardware or software and can instead focus on building their models.

“The speed at which AI systems can be built and deployed on the operational platform is orders of magnitude faster compared to the industry standard tools such as TensorFlow and require far fewer people and decreases the level of technical expertise needed,” Luka Crnkovic-Friis, of Peltarion’s CEO and co-founder, tells me. “All this results in more organizations being able to operationalize AI and focusing on solving problems and creating change.”

In a world where businesses have a plethora of choices, though, why use Peltarion over more established players? “Almost all of our clients are worried about lock-in to any single cloud provider,” Crnkovic-Friis said. “They tend to be fine using storage and compute as they are relatively similar across all the providers and moving to another cloud provider is possible. Equally, they are very wary of the higher-level services that AWS, GCP, Azure, and others provide as it means a complete lock-in.”

Peltarion, of course, argues that its platform doesn’t lock in its users and that other platforms take far more AI expertise to produce commercially viable AI services. The company rightly notes that, outside of the tech giants, most companies still struggle with how to use AI at scale. “They are stuck on the starting blocks, held back by two primary barriers to progress: immature patchwork technology and skills shortage,” said Crnkovic-Friis.

The company will use the new funding to expand its development team and its teams working with its community and partners. It’ll also use the new funding for growth initiatives in the U.S. and other markets.

Lack of transparency in healthcare startups risks another Theranos implosion

Are more Theranos -style scandals looming for investors in healthcare startups?

A team of researchers associated with the Meta-Research Innovation Center at Stanford thinks so. They’ve published a paper warning investors in life sciences startups that a systemic lack of transparency exists in their portfolio companies — creating the possibility for more multi-billion-dollar implosions and scandals like the one that toppled Theranos and its charismatic founder, Elizabeth Holmes.

Indeed, one of the study’s authors, Dr. John Ioannidis, the co-director of the Meta-Research Innovation Center at Stanford and director of the University’s PhD program in Epidemiology and Clinical Research, was  among the first people to identify the risks associated with Theranos and its “stealth research.”

Now Dr. Ioannidis and his co-authors, Ioana A. Cristea and Eli M. Cahan, have published a study surveying the publicly available research from the largest privately held companies in the healthcare space, and found them lacking. 

Most of the highest-valued startups in healthcare have not published any significant scientific literature, the study found. Nearly half of the publications from companies worth more than $1 billion came from only two startups — 23andMe and Adaptive Biotechnologies, according to the paper.

“Many years ago I was the first person to say that Theranos had a problem,” says Ioannidis. “The problem that I had then was that Theranos did not have any peer-reviewed evidence to show.”

In an interview and in their paper, Ioannidis and Cahan warn that investors have overlooked systemic problems created by the lack of transparency among healthcare startups.

They write:

It would be tempting to dismiss the Theranos case as just one rotten apple. However, we worry that the focus on fraud puts aside a more fundamental concern. Fraud is making waves in the news, but stealth research may have a more detrimental impact.

According to the study’s findings, more than half of the healthcare startups that are worth more than $1 billion have published no highly cited papers at all. For companies that were acquired or are publicly traded that number is around 40 percent.

In all, healthcare startups that are currently valued at more than $1 billion published 425 Pubmed papers. And of those papers only 34 (8 percent, including two reviews) were highly cited. For companies with valuations of more than $1 billion that had been acquired or are publicly traded on stock exchanges, the researchers counted 413 papers, of which 47 (11 percent, including nine reviews) were highly cited.

Digging deeper into some of the companies that had high valuations but little or no published research revealed scores of operational and technological issues for the researchers.

For instance, StemCentrx, which was bought for $10.2 billion in 2016 by AbbVie, had published 16 papers — and only one highly cited paper. Since the acquisition, the Food and Drug Administration had imposed a delay on the readout of the company’s phase II trial for its Rova T targeted antibody drug for cancer treatment. In December, a Phase III trial for Rova T as a second-line treatment for patients with advanced small cell lung cancer was halted because the treatment wasn’t working, according to a report in Targeted Oncology.

Acerta Pharma, another healthcare-focused startup focused on cancer treatments, was bought by AstraZeneca for $7.3 billion. That company published nine articles and had one highly cited paper for a very early study of a potential treatment for relapsed chronic lymphocytic leukemia. Acerta received accelerated approval for a drug called acalabrutinib, which treats a rare form of lymphoma called mantle cell lymphoma. Two years ago, AstraZeneca had to retract data and admit that Acerta falsified preclinical data for its drug.

Then there’s Intarcia, the developer of a device for diabetes treatment that’s worth $5.5 billion. That company had its device rejected by the FDA and was forced to lay off staff and halt a couple of later-stage trials. It had only published six papers — none of them very highly cited.

Ultimately, the researchers concluded that highly valued healthcare startups don’t contribute to published research and that the valuation of these companies by investors is divorced from any externally validated data.

For the researchers (and for investors) this should present a problem.

“Many unicorns may be overvalued [21] and subject to unrealistic scientific expectations,” the study’s authors write. And they reject the argument that simply applying for — and receiving — patents is enough to prove that a technology in the healthcare space has been thoroughly vetted. “[Patents] do not offer the same level of documentation as peer-reviewed articles. For example, Theranos had over 100 patents [1], but these were unable to supplant the vacuum in their evidence,” the researchers wrote. 

Even if companies want to protect their technology, there are still ways for them to be more transparent about the results or benefits of their technology. The authors acknowledge that publishing isn’t the primary mission of startups. They can, however publish a few high-value articles, secure their technology through patents and then work with researchers, universities or hospitals to validate the technology and have those organizations publish results of the tests, the authors argue.

As the authors conclude:

Start-ups are key purveyors of innovation and disruption. Consequently, holding them to a minimal standard of evaluation from the scientific community is crucial. Participation in peer review, with all its limitations, is the best way we have to uphold this standard. We are not arguing that start-ups should divert excessive resources to having peer-reviewed papers. However, when their products are destined to affect patient health, they should neither be solely doing marketing. Confidential data sharing with potential investors or regulators cannot replace more open scrutiny by the scientific community.

Pocket’s reading app won’t sound so robotic now

Last year, Mozilla made its first acquisition by snatching up Pocket, the Instapaper competitor that helps you save longer articles for later reading. Today, this popular reading app is getting a major update that gives its app a visual makeover, including a new dark mode, and most importantly, a better way to listen to the content you’ve saved.

Pocket had added a text-to-speech feature several years ago, so you could listen to an audio version of your saved articles, instead of reading them. Instapaper today offers a similar option.

But these text-to-speech engines often sound robotic and mangle words, leading to a poor listening experience. They’ll work in a pinch when you really need to catch up with some reading, and can’t sit down to do it. But they’re definitely not ideal.

Today, Pocket is addressing this problem with the launch of a new listening feature that will allow for a more human-sounding voice. On iOS and Android, the listen feature will be powered by Amazon Polly, Mozilla says.

First introduced at Amazon’s re:Invent developer event in November 2016, Polly uses machine learning technologies to deliver more life-like speech. Polly also understands words in context. For example, it knows that the word “live” would be pronounced differently based on its usage. (E.g. “I live in Seattle” vs. “Live from New York.”) The technology has evolved since to support speech marks, a timbre effect, and dynamic range compression, among other things.

To take advantage of the updated “Listen” feature, users just tap the new icon in the top-left corner of the Pocket mobile app to start playing their articles. It’s like your own personalized podcast, Mozilla notes.

In addition, the app has been given a redesign that gives it a clean, less cluttered look-and-feel, and introduces a new app-wide dark mode and sephia themes, for those who want a different sort of reading experience.

The redesign includes updated typography and fonts, focused on making long reads more comfortable, as well.

[embedded content]

“At Mozilla, we love the web. Sometimes we want to surf, and the Firefox team has been working on ways to surf like an absolute champ with features like Firefox Advance,” said Mark Mayo, Chief Product Officer at Firefox, in a statement about the launch. “Sometimes, though, we want to settle down and read or listen to a few great pages. That’s where Pocket shines, and the new Pocket makes it even easier to enjoy the best of the web when you’re on the go in your own focused and uncluttered space,” he said.

The updated version of Pocket is live on the web, iOS and Android, as of today.