OctoML raises $15M to make optimizing ML models easier

OctoML, a startup founded by the team behind the Apache TVM machine learning compiler stack project, today announced it has raised a $15 million Series A round led by Amplify, with participation from Madrone Ventures, which led its $3.9 million seed round. The core idea behind OctoML and TVM is to use machine learning to optimize machine learning models so they can more efficiently run on different types of hardware.

“There’s been quite a bit of progress in creating machine learning models,” OctoML CEO and University of Washington professor Luis Ceze told me. “But a lot of the pain has moved to once you have a model, how do you actually make good use of it in the edge and in the clouds?”

That’s where the TVM project comes in, which was launched by Ceze and his collaborators at the University of Washington’s Paul G. Allen School of Computer Science & Engineering. It’s now an Apache incubating project and because it’s seen quite a bit of usage and support from major companies like AWS, ARM, Facebook, Google, Intel, Microsoft, Nvidia, Xilinx and others, the team decided to form a commercial venture around it, which became OctoML. Today, even Amazon Alexa’s wake word detection is powered by TVM.

Ceze described TVM as a modern operating system for machine learning models. “A machine learning model is not code, it doesn’t have instructions, it has numbers that describe its statistical modeling,” he said. “There’s quite a few challenges in making it run efficiently on a given hardware platform because there’s literally billions and billions of ways in which you can map a model to specific hardware targets. Picking the right one that performs well is a significant task that typically requires human intuition.”

And that’s where OctoML and its “Octomizer” SaaS product, which it also announced, today come in. Users can upload their model to the service and it will automatically optimize, benchmark and package it for the hardware you specify and in the format you want. For more advanced users, there’s also the option to add the service’s API to their CI/CD pipelines. These optimized models run significantly faster because they can now fully leverage the hardware they run on, but what many businesses will maybe care about even more is that these more efficient models also cost them less to run in the cloud, or that they are able to use cheaper hardware with less performance to get the same results. For some use cases, TVM already results in 80x performance gains.

Currently, the OctoML team consists of about 20 engineers. With this new funding, the company plans to expand its team. Those hires will mostly be engineers, but Ceze also stressed that he wants to hire an evangelist, which makes sense, given the company’s open-source heritage. He also noted that while the Octomizer is a good start, the real goal here is to build a more fully featured MLOps platform. “OctoML’s mission is to build the world’s best platform that automates MLOps,” he said.

You can now buy AWS’ $99 DeepComposer keyboard

AWS today announced that its DeepComposer keyboard is now available for purchase. And no, DeepComposer isn’t a mechanical keyboard for hackers but a small MIDI keyboard for working with the AWS DeepComposer service that uses AI to create songs based on your input.

First announced at AWS re:Invent 2019, the keyboard created a bit of confusion, in part because Amazon’s announcement almost made it seem like a consumer product. DeepComposer, which also works without the actual hardware keyboard, is more of a learning tool, though, and belongs to the same family of AWS hardware like DeepLens and DeepRacer. It’s meant to teach developers about generative adversarial networks, just like DeepLens and DeepRacer also focus on specific machine learning technologies.

Users play a short melody, either using the hardware keyboard or an on-screen one, and the service then automatically generates a backing track based on your choice of musical style. The results I heard at re:Invent last year were a bit uneven (or worse), but that may have improved by now. But this isn’t a tool for creating the next Top 40 song. It’s simply a learning tool. I’m not sure you need the keyboard to get that learning experience out of it, but if you do, you can now head over to Amazon and buy it.

Stocks post worst quarter since 2008 financial crisis

The first quarter of 2020 ended with a whimper — with the Dow Jones Industrial Average, S&P 500 and Nasdaq posting their worst quarter in decades — as the COVID-19 pandemic continues to cause uncertainty and volatility across all major stock market indices.

At the beginning of the quarter, we were still basking in a decade-long bull market. The global pandemic, and the economic havoc it caused, put an end to those halcyon days. All major American indices dropped into bear-market territory March 12, after shedding the requisite 20% from recent highs.

The roller coaster continued, with equities bumping along the bottom, periodically popping up, only to fall again as the epicenter of the pandemic shifted from China to Europe and now the United States. The number of cases in the U.S. has prompted states to issue stay at home orders, putting the brakes on business as usual. As a result, unemployment benefits have skyrocketed. Last week alone, around 3.3 million Americans filed for unemployment benefits, dwarfing numbers set during the 2008-era economic meltdown.

The economic stimulus bill, known as the CARES Act, along with a series of actions taken by the Federal Reserve, have provided some lift. But the volatility continues. For the quarter, Dow Jones is down 24.08%, while the S&P is down 20.67% and Nasdaq is off 15.3%.

Here’s the breakdown of what happened today:

  • Dow Jones Industrial Average: declined 1.85%, or 413.11 to 21,914.37
  • S&P 500: slid 1.61%, or 42.18, to 2,584.47
  • Nasdaq Composite: fell 0.95%, or 74.05, to 7,700.10

All sectors were down today, with the exception of the energy sector, which saw a lift after being battered for weeks. Meanwhile, investors have fled equities for treasuries, pushing yields down. Case in point: U.S. 10-year yields are down 64% in the first quarter.

SaaS shares fell more than most tech equity in today’s trading, with the Bessemer cloud index off a little over 2.5%. The index, which tracks a basket of SaaS and cloud shares, is off around 20% from its recent highs. Shares of modern software companies are therefore still technically in a bear market, though just. If recent gains hold, the index will have made up around 10% of its lost ground since recent lows.

Wrapping on cryptocurrencies as we close the book on the quarter, bitcoin posted a net loss for the period. It’s worth just over $6,400 as we write this post.

What a quarter. What a quarter of surprises and turmoil and cut expectations and downgraded hope. Here’s to a better Q2, if we can manage.

Tech giants should let startups defer cloud payments

Google, Amazon, and Microsoft are the landlords. Amidst the Coronavirus economic crisis, startups need a break from paying rent. They’re in a cash crunch. Revenue has stopped flowing in, capital markets like venture debt are hesitant, and startups and small-to-medium sized businessesf are at risk of either having to lay off huge numbers of employees and/or shut down.

Meanwhile, the tech giants are cash rich. Their success this decade means they’re able to weather the storm for a few months. Their customers cannot.

Cloud infrastructure costs area amongst many startups’ top expenses besides payroll. The option to pay these cloud bills later could save some from going out of business or axing huge parts of their staff. Both would hurt the tech industry, the economy, and the individuals laid off. But most worryingly for the giants, it could destroy their customer base.

The mass layoffs have already begun. Soon we’re sure to start hearing about sizable companies shutting down, upended by COVID-19. But there’s still an opportunity to stop a larger bloodbath from ensuing.

That’s why I have a proposal: cloud relief.

The platform giants should let startups and small businesses defer their cloud infrastructure payments for three to six months until they can pay them back in installments. Amazon AWS, Google Cloud, Microsoft Azure, these companies’ additional infrastructure products, and other platform providers should let customers pause payment until the worst of the first wave of the COVID-19 economic disruption passes. Profitable SAAS providers like Salesforce could give customers an extension too.

There are plenty of altruistic reasons to do this. They have the resources to help businesses in need. We all need to support each other in these tough times. This could protect tons of families. Some of these startups are providing important services to the public and even discounting them, thereby ramping up their bills while decreasing revenue.

Then there are the PR reasons. After years of techlash and anti-trust scrutiny, here’s the chance for the giants to prove their size can be beneficial to the world. Recruiters could use it as a talking point. “We’re the company that helped save Silicon Valley.” There’s an explanation for them squirreling away so much cash: the rainy day has finally arrived.

But the capitalistic truth and the story they could sell to Wall Street is that it’s not good for our business if our customers go out of business. Look at what happened to infrastructure providers in the dotcom crash. When tons of startups vaporized, so did the profits for those selling them hosting and tools. Any government stimulus for businesses would be better spent by them paying employees than paying the cloud companies that aren’t in danger. Saving one future Netflix from shutting down could cover any short-term loss from helping 100 other businesses.

This isn’t a handout. These startups will still owe the money. They’d just be able to pay it a little later, spread out over their monthly bills for a year or so. Once mass shelter-in-place orders subside, businesses can operate at least a little closer to normal, and investors get less cautious, customers will have the cash they need to pay their dues. Plus interest if necessary.

Meanwhile, they’ll be locked in and loyal customers for the foreseeable future. Cloud vendors could gate the deferment to only customers that have been with them for X amount of months or that have already spent Y amount on the platform. The vendors could also offer the deferment on the condition that customers add a year or more to their existing contracts. Founders will remember who gave them the benefit of the doubt.

cloud ice cream cone imagine

Consider it a marketing expense. Platforms often offer discounts or free trials to new customers. Now it’s existing customers that need a reprieve. Instead of airport ads, the giants could spend the money ensuring they’ll still have plenty of developers building atop them by the end of 2020.

Beyond deferred payment, platforms could just push the due date on all outstanding bills to three or six months from now. Alternatively, they could offer a deep discount such as 50% off for three months if they didn’t want to deal with accruing debt and then servicing it. Customers with multi-year contracts could offered the opportunity to downgrade or renegotiate their contracts without penalties. Any of these might require giving sales quota forgiveness to their account executives.

It would likely be far too complicated and risky to accept equity in lieu of cash, a cut of revenue going forward, or to provide loans or credit lines to customers. The clearest and simplest solution is to let startups skip a few payments, then pay more every month later until they clear their debt. When asked for comment or about whether they’re considering payment deferment options, Microsoft declined, and Amazon and Google did not respond.

To be clear, administering payment deferment won’t be simple or free. It could require the giants to change their earnings guidance. Rewriting deals with significantly sized customers will take work on both ends, and there’s a chance of breach of contract disputes. Giants would face the threat of customers recklessly using cloud resources before shutting down or skipping town.

Most taxing would be determining and enforcing the criteria of who’s eligible. The vendors would need to lay out which customers are too big so they don’t accidentally give a cloud-intensive but healthy media company a deferment they don’t need. Businesses that get questionably excluded could make a stink in public. Executing on the plan will require staff when giants are stretched thin trying to handle logistics disruptions, misinformation, and accelerating work-from-home usage.

Still, this is the moment when the fortunate need to lend a hand to the vulnerable. Not a hand out, but a hand up. Companies with billions in cash in their coffers could save those struggling to pay salaries. All the fundraisers and info centers and hackathons are great, but this is how the tech giants can live up to their lofty mission statements.

We all live in the cloud now. Don’t evict us. #CloudRelief

Thanks to Falon Fatemi, Corey Quinn, Ilya Fushman, Jason Kim, Ilya Sukhar, and Michael Campbell for their ideas and feedback on this proposal

Startups are helping cloud infrastructure customers avoid vendor lock-in

For much of the history of enterprise technology, companies tended to buy from a single vendor because it made managing the entire affair much easier while giving them a “single throat to choke” when something went wrong. On the flip side, it also put customers at the mercy of said vendor — and it wasn’t always pretty.

As we move deeper into the cloud model, many IT pros are looking for more flexibility than they had in the past, avoiding the vendor lock-in from the previous generation of enterprise tech, and what being beholden to a single vendor could mean for the bottom line and their own flexibility.

This is something that comes up frequently in discussions about moving workloads from one cloud to another, and is sometimes referred to as a multi-cloud approach. Customers are loath to leave their workloads in the hands of one vendor again and repeat the mistakes of the past. They are looking to have the same flexibility on the infrastructure side that they are getting in the SaaS world, where companies tend to purchase best-of-breed from multiple vendors.

That means, they want the freedom to move workloads between clouds, but that’s not always as easy a prospect as it might seem, and it’s an area where startups could help lead the way.

What’s the problem?

What’s stopping customers from just moving data and applications between clouds? It turns out that there is a complex interlinking of public cloud APIs that help the applications and data work in tandem. If you want to pull out of one public cloud, it’s not a simple matter of just migrating to the next one.

Activist investor Starboard Value taking three Box board seats as involvement deepens

When activist investors Starboard Value took a 7.5% stake in Box last September, there was reasonable speculation that it would begin to try and push an agenda, as activist investors tend to do. While the firm has been quiet to this point, today Box announced that Starboard was adding three members to the 9 member Box board.

At the same time, two long-time Box investors and allies, Rory O’Driscoll from Scale Venture Partners and Josh Stein from DFJ, will be retiring from the board and not seeking reelection at the annual stockholder’s meeting in June.

O’Driscoll involvement with the company dates back a decade, while Stein has been with the company 14 years, from its very earliest days, and has been a big supporter from almost the beginning of the company.

For starters, Jack Lazar, whose credentials including being chief financial officer at GoPro and Atheros Communications, is joining the board immediately. A second new board member from a list to be agreed upon by Box and Starboard will also be joining immediately.

Finally, a third member will be selected by the newly constituted board in June, giving Starboard three friendly votes and the ability to push the Box agenda in a significant way.

At the time it announced it was taking a stake in Box, Starboard telegraphed that it could be doing something like this. Here’s what it had to say in its filing at the time:

“Depending on various factors including, without limitation, the Issuer’s financial position and investment strategy, the price levels of the Shares, conditions in the securities markets and general economic and industry conditions, the Reporting Persons may in the future take such actions with respect to their investment in the Issuer as they deem appropriate including, without limitation, engaging in communications with management and the Board of Directors of the Issuer, engaging in discussions with stockholders of the Issuer or other third parties about the Issuer and the [Starboard’s] investment, including potential business combinations or dispositions involving the Issuer or certain of its businesses, making recommendations or proposals to the Issuer concerning changes to the capitalization, ownership structure, board structure (including board composition), potential business combinations or dispositions involving the Issuer or certain of its businesses, or suggestions for improving the Issuer’s financial and/or operational performance, purchasing additional Shares, selling some or all of their Shares, engaging in short selling of or any hedging or similar transaction with respect to the Shares…”

Box CEO Aaron Levie appeared at TechCrunch Sessions: Enterprise, the week this news about Starboard broke, and he was careful in how he discussed a possible relationship with the firm. “Well I think in their statement actually they really just identified that they think there’s upside in the stock. It’s still very early in the conversations and process, but again we’re super collaborative in these types of situations. We want to work with all of our investors and I think that’ll be the same here,” Levie told us at the time.

Now, the company has no choice, but to work more collaboratively with Starboard as it takes a much more meaningful role on the company board. What impact this will have in the long run is hard to say, but surely significant changes are likely on the way.

Box’s Aaron Levie says it will take creativity and focus to get through this crisis

The COVID-19 virus is touching every aspect of our lives and having a profound impact on individuals, businesses and society at large. Box’s Aaron Levie has built a successful business from dorm room to IPO and beyond. He spoke to TechCrunch today about the level of creativity and focus that it’s going to take to succeed in the current environment.

Levie pointed out that his company was a fledgling startup when the economic downturn hit in 2008, but he thinks this one could have a much greater impact on business than that one did.

“I think Silicon Valley is going to definitely experience this in a very, very significant way. We were building a company in 2008, and that was extremely hard, but I don’t think it is going to compare to how hard the coming year is going to be,” Levie said.

This morning on Twitter, Levie wrote that we are in uncharted territory, and everyone will have to work together to help navigate this crisis.

He believes the government will need to step in to help individuals and businesses alike. “Businesses, who have lots of employees, need to be supported, but fundamentally we need to make sure that we’re focused on all the workers that are out of work, hopefully just temporarily displaced, but we’re going to need a lot of government financial support to get through this,” he said.

For startups, he advised startups to firmly focus on their mission. “It’s about extreme focus right now. It’s about extreme discipline. It’s about making sure that you’re maintaining your culture during this time,” Levie said.

As for his own company, he’s looking a three areas: his employees, his customers and the community. He said his first priority is making sure his employees are safe and healthy and that the hourly workers who support the business normally are being taken care of as we move through this unprecedented situation.

Secondly, he’s making sure that he supports his customers. To that end the company has removed any license limits as customers deal with increased usage with employees working from home.

He has also joined forces with Cloudflare in an effort to provide small businesses with 90 days of free services to help ride out the situation, and he said they would revisit extending these programs if the situation continues.

Thirdly, he says every business who can has to look at ways to support the communities where they live to assist non-profit organizations who are helping in the response. “This is an event where business communities globally are going to have to put more of a concerted effort on this than any issue in modern history,” Levie said.

Levie is not alone in this thinking by any means. He points to other leaders such as Chuck Robbins, Marc Benioff and Tim Cook, all who have stepped up in recent days to offer help and support.

He has built his company from the ground up to one that’s on nearly an $800 million run rate, but like so many business leaders, he is dealing with a situation which, as he said, has no playbook. Like every other CEO, he’s trying to help keep his business thriving, while not losing sight of the needs of the people in his organization, his customers or his community. It’s not an easy balancing act for anyone right now.

Google cancels I/O developer conference in light of COVID-19 crisis

Google announced on Twitter today that it was canceling its annual I/O developer conference out of concern for the health and safety of all involved. It will not be holding any online conference in its place either.

“Out of concern for the health and safety of our developers, employees, and local communities — and in line with recent ‘shelter in place’ orders by the local Bay Area counties — we sadly will not be holding I/O in any capacity this year,” the company tweeted.

This is not a small deal, as Google uses this, and the Google Cloud Next conference, which it has also canceled, to let developers, customers, partners and other interested parties know about what new features, products and services they will be introducing in the coming year.

Without a major venue to announce these new tools, it will be harder for the company to get the word out about them or gain the power of human networking that these conferences provide. All of that is taking a backseat this year over concerns about the virus.

The company made clear that it does not intend to reschedule these events in person or in a virtual capacity at all this year, and will look for other ways to inform the community of changes, updates and new services in the coming months.

“Right now, the most important thing all of us can do is focus our attention on helping people with the new challenges we all face. Please know that we remain committed to finding other ways to share platform updates with you through our developer blogs and community forums,” the company wrote.

Big opening for startups that help move entrenched on-prem workloads to the cloud

AWS CEO Andy Jassy showed signs of frustration at his AWS re:Invent keynote address in December.

Customers weren’t moving to the cloud nearly fast enough for his taste, and he prodded them to move along. Some of their hesitation, as Jassy pointed out, was due to institutional inertia, but some of it also was due to a technology problem related to getting entrenched, on-prem workloads to the cloud.

When a challenge of this magnitude presents itself and you have the head of the world’s largest cloud infrastructure vendor imploring customers to move faster, you can be sure any number of players will start paying attention.

Sure enough, cloud infrastructure vendors (ISVs) have developed new migration solutions to help break that big data logjam. Large ISVs like Accenture and Deloitte are also happy to help your company deal with migration issues, but this opportunity also offers a big opening for startups aiming to solve the hard problems associated with moving certain workloads to the cloud.

Think about problems like getting data off of a mainframe and into the cloud or moving an on-prem data warehouse. We spoke to a number of experts to figure out where this migration market is going and if the future looks bright for cloud-migration startups.

Cloud-migration blues

It’s hard to nail down exactly the percentage of workloads that have been moved to the cloud at this point, but most experts agree there’s still a great deal of growth ahead. Some of the more optimistic projections have pegged it at around 20%, with the U.S. far ahead of the rest of the world.

Spectro Cloud launches with $7.5M investment to help developers build Kubernetes clusters their way

By now, we know that Kubernetes is a wildly popular container management platform, but if you want to use it, you pretty much have to choose between having someone manage it for you or building it yourself. Spectro Cloud emerged from stealth today with a $7.5 million investment to give you a third choice which falls somewhere in the middle.

The funding was led by Sierra Ventures with participation from Boldstart Ventures.

Ed Sim, founder at Boldstart says he liked the team and the tech. “Spectro Cloud is solving a massive pain that every large enterprise is struggling with; how to roll your own Kubernetes service on a managed platform without being beholden to any large vendor.” Sim told TechCrunch.

Spectro co-founder and CEO Tenry Fu says that an enterprise should not have to compromise between control and ease of use. “We want to be the first company that brings an easy-to-use managed Kubernetes experience to the enterprise, but also gives them the flexibility to define their own Kubernetes infrastructure stacks at scale,” Fu explained.

Fu says that the stack in this instance consists of the base operating system to the Kubernetes version to the storage, networking and other layers like security, logging, monitoring, load balancing or anything that’s infrastructure related around Kubernetes.

“Within an organization in the enterprise you can serve the needs of your various groups, down to pretty granular level with respect to what’s in your infrastructure stack, and then you don’t have to worry about lifecycle management,” he explained. That’s because they handle that for you, while still giving you that control.

That not only gives enterprise developers greater deployment flexibility, it gives them the ability to move between cloud infrastructure providers more easily, something that is top of mind today as companies don’t want to be locked into a single vendor.

“There’s an infrastructure control continuum that forces enterprises into trade offs against these needs. At one extreme, the managed offerings offer a kind of nirvana around ease of use, but it’s at the expense of control over things like the cloud that you’re on or when you adopt new ecosystem options like updated versions of Kubernetes.”

Fu and his co-founders have a deep background in this, having previously been part of CliQr, a company that helped customers manage applications across hybrid cloud environments. They sold that company to Cisco in 2016, and began developing Spectro Cloud last spring.

It’s early days, but the company has been working with 16 Beta customers.