Technology

San Francisco-based startup Particle was one of the rising stars in the Internet of Things space, raising more than $81 million to date on the promise of helping to manage and secure the next generation of connected devices.

But the company is only now emerging from what its co-founder and chief executive Zach Supalla called a &turbulent period,& prompting layoffs and cost-cutting to help stay afloat, TechCrunch has learned.

Founded in 2012, Particle snagged $40 million in its Series C fundraise last October from big industrial investors, including Qualcomm Ventures and Energy Impact Partners, signaling strong support for the companymission. The startup pitches its flagship platform as an all-in-one solution to manage and secure IoT devices with encryption and security, but also scalability and data autonomy.

But a recent email sent by Supalla to his staff — obtained by TechCrunch — shows the company is course-correcting after a recent revenue miss.

The email, which the company confirmed was sent by the chief executive, said Particle laid off 14 staff members earlier this month, representing about 10% of the company. The layoffs of both engineering and support staff came just weeks after co-founder and chief technology officer Zachary Crockett quietly departed the company for &unrelated& reasons, said Supalla. (Crockett did not respond to a request for comment.)

According to Supallaemail to staff, Particlerevenue goal in 2019 was $16 million, but it ended the year with $10.3 million. Supalla cited, among other things, &operational challenges& with the business that he said kept the company &from executing as well as we could.&

Supalla said the company still has a &flush& bank account with more than $30 million in the bank, but the companycurrent burn rate of $2 million per month is &uncomfortably high.&

&We would only have until early 2021 to prepare for the next stage of financing the company,& he said.

The email added that the company is bringing on $10 million in venture debt, but Supalla told TechCrunch that the deal is &still in progress.& Particle is aiming to reduce its burn rate to about $1.6 million per month, which Supallaemail said would be achievable with the recent layoffs and reducing discretionary budgets, including marketing.

The cost-cutting will &put us in a position of financial strength,& the email said, adding that the company has &no intentions& of further layoffs.

Although the 14 employees have been given severance, one source said that some are still waiting for the payouts — some two weeks after the announcement — which Supalla confirmed in an email. TechCrunch also learned that former staff were asked to sign non-disclosure agreements. Supalla told TechCrunch that these agreements come with non-disparagement clauses, but that anyone laid off that wanted to be released from the non-disparagement terms would be.

Supallaemail is hardly the death knell for the company, but questions remain about its revenue targets and its efforts to reduce its monthly burn rate. The chief executiveemail said, candidly, that while layoffs can signal financial duress, they&re all too often made too late and &as a last resort.&

&Thatnot whathappening here,& said Supalla. &We have plenty of money in the bank and are making prudent cuts to strengthen the business.&

Got a tip? You can send tips securely over Signal and WhatsApp to +1 646-755&8849.

Cartesiam, a startup that aims to bring machine learning to edge devices powered by microcontrollers, has launched a new tool for developers who want an easier way to build services for these devices. The new NanoEdge AI Studio is the first IDE specifically designed for enabling machine learning and inferencing on Arm Cortex-M microcontrollers, which power billions of devices already.

As Cartesiam GM Marc Dupaquier, who co-founded the company in 2016, told me, the company works very closely with Arm, given that both have a vested interest in having developers create new features for these devices. He noted that while the first wave of IoT was all about sending data to the cloud, that has now shifted and most companies now want to limit the amount of data they send out and do a lot more on the device itself. And thatpretty much one of the founding theses of Cartesiam. &Itjust absurd to send all this data — which, by the way, also exposes the device from a security standpoint,& he said. &What if we could do it much closer to the device itself?&

The company first bet on Intelshort-lived Curie SoC platform. That obviously didn&t work out all that well, given that Intel axed support for Curie in 2017. Since then, Cartesiam has focused on the Cortex-M platform, which worked out for the better, given how ubiquitous it has become. Since we&re talking about low-powered microcontrollers, though, itworth noting that we&re not talking about face recognition or natural language understanding here. Instead, using machine learning on these devices is more about making objects a little bit smarter and, especially in an industrial use case, detecting abnormalities or figuring out when ittime to do preventive maintenance.

Today, Cartesiam already works with many large corporations that build Cortex-M-based devices. The NanoEdge Studio makes this development work far easier, though. &Developing a smart object must be simple, rapid and affordable — and today, it is not, so we are trying to change it,& said Dupaquier. But the company isn&t trying to pitch its product to data scientists, he stressed. &Our target is not the data scientists. We are actually not smart enough for that. But we are unbelievably smart for the embedded designer. We will resolve 99% of their problems.& He argues that Cartesiam reduced time to market by a factor of 20 to 50, &because you can get your solution running in days, not in multiple years.&

One nifty feature of the NanoEdge Studio is that it automatically tries to find the best algorithm for a given combination of sensors and use cases and the libraries it generates are extremely small and use somewhere between 4K to 16K of RAM.

NanoEdge Studio for both Windows and Linux is now generally available. Pricing starts at €690/month for a single user or €2,490/month for teams.

Eagle-eyed readers will recall that we mentioned M1 Finance earlier today in our look at a few trends in the fintech industry. We&re back with the firm this afternoon as it has a bit of news thatworth discussing.

Chicago-based M1 Finance announced today that it has reached the $1 billion assets under management mark, or AUM. Reaching AUM thresholds provides useful milestones that we can use to track the progress of various players in the fintech and finservices worlds.

M1 is an interesting company, bringing together a number of products to form a single platform. Its hybrid nature makes comparing its AUM to other companies& histories a bit dicey. Still, for reference, Wealthfront, a roboadvisor, announced that it started 2013 with AUM of $100 million, and closed that year with $538 million. By mid-2014, Wealthfront had $1 billion AUM. Today it has over $20 billion.

So, the numbers matter, and reaching thresholds can help us understand where a company is in its maturity cycle.

Lettalk about M1 Finance AUM growth, its revenue growth and its product model. Ita neat company with a history of efficient growth.

Growth, product

We&ll start with product, as how the company approaches its feature-set helps explain how the service is priced, which in turn helps us grok the companygrowth.

M1 is not a roboadvisor, or a simple neobank, or a lending product; itall three at once, providing effectively the digital equivalent of a full-service bank, admittedly in the form of an online experience instead of a brick-and-mortar outlet. M1 users can open investment accounts, checking accounts, get a debit card and borrow money against their investment portfolios; ita cohesive feature set.

And one that lets M1 price its products lower as a group than it could individually. During a call with M1CEO Brian Barnes about the companyAUM milestone, the executive connected the companylong-term vision to its ability to price aggressively. (All fintechs are expanding their platforms, itworth noting, meaning that, in time, nearly every fintech player will offer an array of services; Wealthfront, famous for its work in roboadvising, now also offers savings and borrowing capabilities.)

Barnes said that M1 has long wanted to &manage the bulk of [its users&] financial assets, not create a sort of low-friction acquisition hook& to bring in smaller-dollar accounts. This, in turn, means that M1 can have higher per-user sums on its books, which, it appears, helped the company reduce prices on a per-product basis.

HereBarnes connecting per-account totals to pricing:

Managing more of someonefinancial assets, and financial life, is going to be more economical. What it allows us to do is maintain lower margins per product, but have enough margin on the entire financial relationship that we can build a very sustainable durable, long-lasting business.

Thatneat! And folks with lots of money expect low fees, especially in the Robinhood-era, so the setup probably helps with attracting users.

Revenue

Summing so far, M1 runs a broad set of financial products, attracting more dollars-per-user than other companies, perhaps, which lets it charge, in its view, lower prices.

How low? Barnes told TechCrunch that his company is &building [its] business model to make 1% of assets we manage [into] top line. So every billion bucks on the platform will be 10 million dollars in recurring revenue. And it is a relatively linear relationship.& The CEO later extended the point, saying that when his firm has $10 billion in AUM, it will generate $100 million.

This means that as M1 scales, we&ll be able to know with reasonable confidence how much revenue itdriving.

The company charges in the manner you&d expect, with incomes from loaning money, interchange and a SaaS-product called M1 Plus that lowers some fees and provides interest on checking accounts, costing $125 yearly.

Now that M1 is big enough to matter, it has to double, and then double again. We&ll know how well thatgoing based on how quickly the company reaches the $2 billion mark.

Last November, Twitter rolled out its Hide Replies feature to all users worldwide. The feature, largely designed to lessen the power of online trolls to disrupt conversations, lets users decide which replies to their tweets are placed behind an extra click. Today, Twitter is making Hide Replies available to its developer community, allowing for the creation of tools that help people hide the replies to their tweets faster and more efficiently, says Twitter.

These sorts of tools will be of particular interest to businesses and brands who maintain a Twitter presence, but whose accounts often get too many replies to tweets to properly manage on an individual basis. With Hide Replies now available as a new API endpoint, developers can create tools that automatically hide disruptive tweets based on factors important to their customers — like tweets that include certain prohibited keywords or those that score high for being toxic, for example.

Ahead of todaylaunch, Twitter worked with a small number of developers who are now releasing tools that take advantage of the added functionality.

Jigsaw, an Alphabet-owned company tackling the worst of the web, has integrated Twitternew endpoint with its Perspective API, which uses A.I. to score tweets based on their toxicity. The integration will automatically hide replies that exceed a certain toxic threshold (.94), freeing up the time it would otherwise take to comb through replies manually.

A scripting platform for business workflows, reshuffle, has used the endpoint to develop scripts that detect and hide replies based on keywords or even by user.

Dara Oladosu, the creator of the popular app QuotedReplies, also used the endpoint to build a new app called Hide Unwanted Replies. The app today automatically hides replies by keywords or Twitter handles. Soon, it will add support for hiding replies from likely troll or bot accounts — including tweets from user accounts created too recently or from accounts with few followers.

Hide Replies has been one of Twittermore controversial launches to date, as it could potentially allow users to silence critics or stifle dissent even when warranted — such as in the case of refuting misinformation or propaganda, for example. Others argue itnot really helping address online abuse; the abuse still occurs, but in the shadows. One organization even recently leveraged Hidden Replies for a clever online campaign about how domestic violence goes unseen which further illustrates this problem.

Nevertheless, adoption of Hide Replies is growing, with organizations like the CIA even leveraging it on some tweets.

The new Twitter API endpoint for Hide Replies is available today to all developers in a production-ready form, Twitter says, initially through Twitter Developer Labs. This program launched last year to serve as a way for developers to try out Twitterlatest APIs ahead of their wider release and offer feedback. Participation in Twitter Developer Labs is free, but interested developers have to sign up using an approved developer account. Twitter is also inviting developers building with the new endpoint to collaborate with the company by way of the community forums.

Based on early feedback from the first testers, Twitter says italready making a few changes to the endpoint including support to unhide replies via the endpoint, a higher rate limit to support high-volume use cases, and a way to retrieve a list of replies that indicate if they&re hidden or not.

Up, up, down, down, left, right, left, right, B, A, then start. Sound familiar? The Konami Code, as this sequence came to be known, is one of the most recognizable artifacts of an earlier era of gaming. Kazuhisa Hashimoto, its creator, has used up the last of his 30 lives.

Hashimoto was a programmer at Konami, and created the code during the development of one of Konamibest-known games of the 8-bit era: Gradius. Anyone who played it will remember the crushing difficulty of this iconic side-scrolling shooter.

Even the developers, it turns out, found it a bit of a hassle to get through repeatedly for testing purposes. Thatwhy during the porting process from arcade to NES, Hashimoto made himself a bit of a shortcut to make things a bit easier for himself.

He created a special command that would award the player the most crucial items for surviving the gamechallenges. The sequence to activate it needed to be easy for him to remember during his many playthroughs, but extremely unlikely for a player to input by accident. And so he settled on the well-known &up, up, down, down, left, right, left, right, B, A& — after which is usually appended &start,& since the code is often entered while paused or at the title screen.

Fate intervened here and the code, which was meant to be removed before the team wrapped up, was forgotten about and ended up in the shipping product. Somehow word got out about the code (who knows how such things transpired in the &80s — probably it was published in Nintendo Power) and, given the extreme difficulty of the game, its necessity led to the code being adopted by pretty much everyone who bought Gradius.

And so millions of kids who grew up in the &80s and &90s learned the Konami Code by heart, though its effects differed from game to game, it generally made things considerably easier. For instance, in the infamous (and still amazing) Contra for NES, the code gives the player 30 lives, which honestly is about the bare minimum necessary to complete that brutal game.

The code persisted for many years and across generations, though it also began to mutate — in Gradius III for SNES the code caused the playership to self-destruct, as if telling them that cheaters never prosper. Even games by other publishers used the code, as a joke or in earnest.

Soon the Konami Code was a staple of geek culture. I myself owned a shirt with the code on it, and listened to a band by that name. It showed up in TV, movies, anywhere an &80s kid had a chance to slip it in. Even if it wasn&t used by name or with the exact sequence, the code became shorthand for all other cheat codes. Hashimoto had unknowingly created a proto-meme that infiltrated gaming culture worldwide, becoming one of the most widely recognizable aspects of it for decades to come.

All because he found his own game too hard to play.

Those were the days when development teams were on the order of 10-20 people, and the choices of a single person could change everything. These days a cheat code would probably have to be approved and playtested during alpha and beta, and shared with strategic partners for the printed strategy guides well ahead of release.

Hashimotocontribution to the gaming world was an accident, but on no account does that downplay his or the codeimportance. He represented the lasting power of an earlier era of gaming and game creation, and accident or not, his legacy is a powerful one.

TechCrunch has learned that $28 million-funded crypto startup Tagomi will be the newest member of the Libra Association that governs the Facebook-backed Libra stablecoin. A formal announcement is slated for Friday or next week.

Tagomi offers a platform that helps large traders and funds easily access cryptocurrency markets. The news comes days after Libra added Shopify, a reversal of dwindling membership after major partners like Visa, PayPal and Stripe dropped out late last year.

We&ve reached out to the Libra Association and have been promised a response by Facebookcommunications team.

Joining Libra means Tagomi will be expected to contribute at least $10 million toward developing the cryptocurrency, with that investment eligible to reap dividends from interest earned on money kept in the Libra Reserve. Tagomi will also operate a node that validates transactions coming through the Libra blockchain.

Tagomi was founded by Jennifer Campbell, a former investor at Union Square Ventures, which is also a Libra Association Member. The company has 25 employees across five offices. Tagomi will be the 22nd member of the Libra Association, according to information from the startuppress representative, who was apparently supposed to hold this news until later. &Tagomi is joining the Libra Foundation and Jennifer will be the newest member,& they emailed TechCrunch. We&ll update this story following our interview with Campbell tomorrow.

Campbell and Tagomi will offer technical and policy support to Libra in an effort to make the cryptocurrency more safe and compliant with international law. That will be critical for the Libra Association to get the green light from regulators for a launch in 2020 like it originally planned. Lawmakers in the U.S. and EU have slammed Libra in hearings and the press over its potential to facilitate money laundering, harm privacy and destabilize the global financial system.

The full membership of the Libra Association is now:

Current Members: