Google Cloud vets launch Seattle startup Kaskada to bolster machine learning tech with real-time data

Google Cloud vets launch Seattle startup Kaskada to bolster machine learning tech with real-time data

2:28pm, 2nd August, 2019
The Kaskada leadership team, from left to right: Davor Bonaci, Ben Chambers, and Emily Kruger. (Kaskada Photo) After spending several years working at Google Cloud, and saw an opportunity to help companies take better advantage of machine learning technology. Their idea turned into , a Seattle-based startup that is launching out of stealth mode and unveiling its software that uses real-time, event-based data to bolster machine learning features. Davor Bonaci. (Kaskada Photo) More and more companies are implementing machine learning capabilities into their workflows to serve up better recommendations, detect fraud, and other related applications that use the burgeoning technology. But Kaskada contends that these models aren’t using the most up-to-date information, resulting in stale data and poor predictions that don’t accurately reflect the needs of a given user. The startup’s tools let companies implement machine learning features that fully take advantage of up-to-date streaming data. “There is lots of evidence that this is not done as well as it could be done,” Bonaci said of using real-time data. “Companies are leaving money on the table.” Kaskada has raised $1.8 million from investors including Voyager Capital; NextGen Venture Partners; Founders’ Co-op; and Bessemer Venture Partners. The company, founded in January 2018, employs four people and expects to grow. In March it hired , a veteran of Amazon Web Services, as vice president of product. We caught up with Bonaci for this , a regular GeekWire feature. Continue reading for his answers to our questionnaire. What does your company do? Kaskada is a machine learning studio that uses event-based data to compute feature vectors for machine learning in real time. Kaskada empowers data scientists by allowing them to discover, test, and deploy features from event-based data sources in a collaborative, version-controlled environment. By empowering data scientists we help organizations make better predictions and drive more impact from machine learning. Inspiration hit us when: All the time — we’re inspired by progress. Every conversation with data scientists and data leaders helps us refine our vision and make a better, more impactful product. VC, Angel, or Bootstrap: VC. We’ve been incredibly lucky with our investors so far, which include Voyager Capital, NextGen Venture Partners, Founders’ Co-op, and Bessemer Venture Partners. We are also supported by a group of angels that includes directors and senior vice presidents of companies like Google, Twitter and Yelp. Not only have they provided the working capital, but they are also meaningfully helping build the company. Their insight, personal networks, and day-to-day support have been instrumental in getting where we are today. The value we have gotten from our investors is as important — if not more important — than the funding itself. Our ‘secret sauce’ is: Streaming data of course! Our team has deep experience in building distributed systems for data streams and data processing and believe we can fundamentally change how ML is practiced by helping companies harness the power of real-time data. The smartest move we’ve made so far: We came to the startup world with a lot of experience in the data space which also meant we had many existing opinions and biases about it. It can be hard to listen carefully, probe, and ask the right questions if you think you already know the answer. It was important for us to forget what we thought we knew and look at the space with fresh eyes. We also had to be willing to admit when we were wrong and refocus our direction based on what we heard from customers. Putting the customer stories first allowed us to learn and ultimately make much better decisions about product and company direction than we would have made in a vacuum. The biggest mistake we’ve made so far: Gauging time it will take to get to major milestones. Everything takes longer than you expect that it will — particularly if you’re an optimistic person! Sometimes those same delays can end up ultimately being positive, though, as you realize a much better way of achieving the same goal. Which leading entrepreneur would you most want working in your corner? Success doesn’t depend on a single individual. We believe that building a strong team that can work together toward a common vision is more important than any single individual. Our favorite team building activity is: Game night! We have a weekly team game night and (optional) whiskey tasting. We typically play various cooperative board games, which makes it more about winning together. Our current favorite is Hanabi. The biggest thing we look for when hiring is: Culture fit. Building a company is a journey requiring significant growth — both personally and as a group. We’re looking for people who want to be part of that journey and actively participate in that growth. We’re looking for people who would have fun participating in lively discussions as we seek to push each other and the company to be the best we can be. What’s the one piece of advice you’d give to other entrepreneurs just starting out: Pick your team and supporters wisely. They will make you or break you. No other early decision is more important than that one. When you start a new company, there are many people seeking to be involved. Regardless of the role, you’ll hear how much they can help you. But, there are no shortcuts; you and your team will have to solve the hard problems. Always focus on the team and the people who are committed to the long-term success of the company.
Experts from Google, T-Mobile and other tech frontiers weigh in on the future of AI

Experts from Google, T-Mobile and other tech frontiers weigh in on the future of AI

11:30pm, 25th April, 2019
SalesPal CEO Ashvin Naik, Google Cloud’s Chanchal Chatterjee, Audioburst’s Rachel Batish and T-Mobile’s Chip Reno discuss the future of artificial intelligence at the Global AI Conference in Seattle. (GeekWire Photo / Alan Boyle) Artificial intelligence can rev up recommendation engines and make self-driving cars safer. It can even . But what else will it be able to do? At today’s session of the , a panel of techies took a look at the state of AI applications — and glimpsed into their crystal balls to speculate about the future of artificial intelligence. The panelists included Chanchal Chatterjee, AI leader at ; Ashvin Naik, CEO of , which markets AI-enabled sales analysis tools; Rachel Batish, vice president of product for , an audio indexing service; and Chip Reno, senior advanced analytics manager at . The moderator was Shailesh Manjrekar, head of product and solutions marketing for , a multi-cloud data storage and management company. Here are five AI frontiers that came up in today’s conversations, plus a couple of caveats to keep in mind: Smarter grocery stores: AI-enabled grocery shopping was pioneered right here in Seattle at , but the trend is catching on. Today called the Intelligent Retail Lab in Levittown, N.Y. Britain’s takes a different tack: Users fill up a virtual shopping cart, then schedule a one-hour delivery slot. Google Cloud helped Ocado develop the , including a recommendation engine that figures out customers’ shifting preferences, an algorithm that handles and prioritizes customer service emails, and a as Ocado’s previous system. Energy-saving server farms: Chatterjee pointed to how Google used its DeepMind machine learning platform to . Before AI was put on the case, 10 years’ worth of efficiency measures could reduce energy usage by merely 12 percent, he said. Within six months, AI brought about a 40 percent reduction. “That was a huge difference that AI made in a very short amount of time that we could not do with 10 years of research,” Chatterjee said. Financial market prediction: Hedge fund managers and bankers are already , detect market manipulation and assess credit risks. But Chatterjee said the models are getting increasingly sophisticated. AI is being used to predict how margin trades could play out, or whether undervalued financial assets are ripe for the picking. AI models could even anticipate . “When the lock-in period expires … that’s a great time to short,” Chatterjee said. Deeper, wider AI conversations: Chatterjee predicted that our conversations with voice assistants are likely to get wider, deeper and more personal as AI assistants become smarter. Audioburst’s Batish said conversational AI could provide a wider opening for smaller-scale startups and for women in tech. “Women are very much prominent in conversational applications and businesses,” she said. Salespal’s Naik agreed with that view — but he worried about the dearth of compelling applications, based on his own company’s experience with voice-enabled devices like Amazon Echo and Google Home. “They’re gathering dust. … We use them just to listen to music or set up alarms. That’s it,” he said. AI for good, or evil? Chatterjee said AI could be a powerful tool to root out fraud and corruption. AI applications could be built “to see what influence relationships have on outcomes — that tells you if there are any side deals being made,” he said. But Batish worried about the rise of , virtual and . “I’m actually afraid of what that could bring into our world,” she said. “It would be interesting to see how companies are trying to be able to monitor or identify fake situations that are being built out of very complicated AI.” Watch out for job disruption: Many studies have pointed out that automation is likely to disrupt employment sectors, especially in the service, manufacturing and transportation sectors. “Anything that is repetitive, that can be extracted from multiple sources, that doesn’t have a lot of creativity amd innovation, is at risk due to AI,” Chatterjee said. “That means that more people will have to move into other sectors.” Watch out for the hype: “I’d like to see people get away from the hype a little bit,” T-Mobile’s Reno said. “I’m on the client side, so I see all the pitches involving AI and ML or deep learning. … A lot of times, AI is not applicable to certain use cases where we’re applying it. Just good old-fashioned statistics or business intelligence is fine. So I think that the future of AI relies on getting past the hype and getting more into aligning these awesome tools and algorithms to specific business cases.”
Spawned by Google, Wing wins FAA’s air carrier certification for drone deliveries

Spawned by Google, Wing wins FAA’s air carrier certification for drone deliveries

12:10pm, 23rd April, 2019
Wing’s drone makes a delivery. (Wing Photo) Alphabet’s has stolen a march on Amazon’s plans for drone domination by winning air carrier certification from the Federal Aviation Administration. “Air Carrier Certification means that we can begin a commercial service delivering goods from local businesses to homes in the United States,” celebrating the milestone. Wing was from (formerly known as Google X), and has been taking part in an in Southwest Virginia. The company has also conducted a test program in Australia that involved more than 3,000 drone deliveries to doorsteps, backyards and driveways. In all, Wing’s drones have flown more than 70,000 test flights, and is starting up . Wing said the data submitted to the FAA for certification showed that “a delivery by wing carries a lower risk to pedestrians than the same trip made by car.” Transportation Secretary Elaine Chao hailed the certification. “This is an important step forward for the safe testing and integration of drones into our economy. Safety continues to be our No. 1 priority as this technology continues to develop and realize its full potential,” she said in a statement. Wing said its next step will be to further its participation in the Virginia pilot program. “For the next several months, we’ll be reaching out to businesses and community members in the Blacksburg and Christiansburg areas to demonstrate our technology, answer questions, and solicit feedback with the goal of launching a delivery trial later this year,” the company said. Amazon has been conducting its own drone delivery test flights in locations ranging from Israel and France to . The Seattle-based online retailing giant showed off more than two years ago. Amazon missed out on participating in the FAA’s first wave of drone pilot programs, however. We’ve reached out to Amazon for comment, and will update this item with anything we hear back.
Google Home’s Philips Hue integration can now wake you up gently

Google Home’s Philips Hue integration can now wake you up gently

12:23pm, 17th April, 2019
Maybe you love the sound of your alarm clock blaring in the morning, heralding a new day full of joy and adventure. More likely, though, you don’t. If you prefer a more gentle wake-up (and have invested in some smart home technology), here’s some good news: Home lets you use your lights to wake you up by slowly changing the light in your room. first this integration at CES earlier this year, with a planned rollout in March. Looks like that took a little while longer, as Google and Philips gently brought this feature to life. Just like you can use your Home to turn on “Gentle Wake,” which starts changing your lights 30 minutes before your wake-up time to mimic a sunrise, you also can go the opposite way and have the lights mimic sunset as you get ready to go to bed. You can either trigger these light changes through an alarm or with a command that starts them immediately. While the price of white Hue bulbs has come down in recent years, colored hue lights remain rather pricey, with single bulbs going for around $40. If that doesn’t hold you back, though, the Gentle Sleep and Wake features are now available in the U.S., U.K., Canada, Australia, Singapore and India in English only.
Google is creating its own first-party game studio

Google is creating its own first-party game studio

1:33pm, 19th March, 2019
just unveiled at a conference in San Francisco, its cloud gaming platform. While most of the conference showcased well-known games you can play on your PC, Xbox One or Playstation 4, the company also announced that it is launching its own first-party game studio, Stadia Games and Entertainment. Jade Raymond is going to head the studio and was here to announce the first details. The company is going to work on exclusive games for Stadia. But the studio will have a bigger role than that. "I'm excited to announce that, as the head of Stadia Games and Entertainment, I will not only be bringing first party game studios to reimagine the next generation of games,” Raymond said. “Our team will also be working with external developers to bring all of the bleeding edge Google technology you have seen today available to partner studios big and small." Raymond has been working in the video game industry for more than 15 years. In particular, she was a producer for Ubisoft in Montreal during the early days of the Assassin’s Creed franchise. She also worked on Watch Dogs before leaving Ubisoft for Electronic Arts. She formed Motive Studios for Electronic Arts and worked with Visceral Games, another Electronic Arts game studio. She was working on an untitled single player Star Wars video game, but Visceral Games closed in 2017 and the project has been since then. According to Google, 100 studios around the world have already received development hardware for Stadia. There are over 1,000 engineers and creatives working on Stadia games or ports right now. Stadia uses and a Linux operating system. Games that are already compatible with Linux should be easy to port to Stadia. But there might be more work for studios focused on Windows games. According to , the cloud instance runs on Debian and features Vulkan. The machine runs an x86 CPU with a “custom AMD GPU with HBM2 memory and 56 compute units capable of 10.7 teraflops”. That sounds a lot like the AMD Radeon RX Vega 56, a relatively powerful GPU but something not as powerful as what you can find in high-end gaming PCs today. Google will be running a program called Stadia Partners to help third-party developers understand this new platform.
Google makes it easier for cheap phones and smart devices to encrypt your data

Google makes it easier for cheap phones and smart devices to encrypt your data

9:20pm, 7th February, 2019
Encryption is an important part of the whole securing-your-data package, but it’s easy to underestimate the amount of complexity it adds to any service or device. One part of that is the amount of processing encryption takes — an amount that could be impractical on small or low-end devices. wants to change that with a highly efficient new method called Adiantum. Here’s the problem. While encryption is in a way just transforming one block of data reversibly into another, that process is actually pretty complicated. Math needs to be done, data read and written and reread and rewritten and confirmed and hashed. For a text message that’s not so hard. But if you have to do the same thing as you store or retrieve megabyte after megabyte of data, for instance with images or video, that extra computation adds up quick. Lots of modern smartphones and other gadgets are equipped with a special chip that performs some of the most common encryption algorithms and processes (namely AES), just like we have GPUs to handle graphics calculations in games and such. But what about older phones, or cheaper ones, or tiny smart home gadgets that don’t have room for that kind of thing on their boards? Just like they can’t run the latest games, they might not be able to efficiently run the latest cryptographic processes. They can still encrypt things, of course, but it might take too long for certain apps to work, or drain the battery. Google, clearly interested in keeping cheap phones competitive, is tackling this problem by creating a special encryption method just for low-power phones. They call it Adiantum, and it will be optionally part of Android distributions going forward. , but the gist is this. Instead of using AES it relies on a cipher called ChaCha. This cipher method is highly optimized for basic binary operations, which any processor can execute quickly, though of course it will be outstripped by specialized hardware and drivers. It’s well documented and already in use lots of places — this isn’t some no-name bargain bin code. As they show, it performs way better on earlier chipsets like the Cortex A7. The Adiantum process doesn’t increase or decrease the size of the payload (for instance by padding it or by appending some header or footer data), meaning the same number of bytes come in as go out. That’s nice when you’re a file system and don’t want to have to set aside too many special blocks for encryption metadata and the like. Naturally new encryption techniques are viewed with some skepticism by security professionals, for whom the greatest pleasure in life is to prove one is compromised or unreliable. Adiantum’s engineers say they have “high confidence in its security,” with the assumption (currently reasonable) that its component “primitives” ChaCha and AES are themselves secure. We’ll soon see! In the meantime don’t expect any instant gains, but future low-power devices may offer better security without having to use more expensive components — you won’t have to do a thing, either. Oh, and in case you were wondering: Adiantum is named after the genus of the maidenhair fern, which in the Victorian language of flowers (floriography) represents sincerity and discretion.