& meet dozens of singles today!
User blogs
Making and keeping the web accessible is a full-time job, and like any other development role, accessibility tools need to evolve to keep up with the times. Evinced is a startup that promises both richer and faster checks of websites in production or in progress, and it just raised $17M to take its tools to the next level.
Because accessibility problems can happy in so many ways, it often takes a lot of manual code review to catch the errors. Even a team thinking about making their site fully accessible from the start — which should be everyone — can miss that this script doesn’t hook into that variable right if this menu is opened, and so on.
There’s automated code review, but it can be slow and bulky. Evinced is making a powerful, streamlined tool that checks a website in a fraction of a second while you’re using it, presenting the problems in a way that’s easy for devs to share and address. It also doesn’t trip up on the fancy, javascript-heavy web apps that millions use today.
Here’s an example of a modern website that looks fine but is obviously (for demo purposes) riddled with accessibility issues. The video gives a good breakdown of what this part of the Evinced product does:
Honestly, that’s how it feels like it ought to look, but existing enterprise-level tools probably aren’t quite so efficient. And as you can see, the tool responds instantly while the user (that is to say, the developer) proceeds through the various actions the site enables. It could be, after all, that auditing the site before anyone fills in a form or pulls down any menus could give a misleading green light.
The inspector also brings in a bit of AI in the form of smart rules and computer vision, so if an element looks like a menu or button but isn’t labeled correctly, it isn’t fooled. Those elements do have distinct styles and roles: if something can be clicked and turns into a list that the user chooses from, well, it’s a pulldown menu whether it’s called that or not.
Image Credits: Evinced
Naturally there are also quick fixes suggested and the ability to easily export the issues for formal inspection by the boss, as well as other expected features for a web development tool. It’s available as a Chrome extension, or as an API or automated part of other analysis or commit actions, throwing its list of errors in with the rest.
The company formed back in 2018, when they started development. The next year they hooked up with a few large enterprises to see about integrating and testing within their ecosystems. Capitol One became their biggest customer and is now an investor.
“We have since deployed our products in production at Capital One (means they are used every day – and power their end-to-end accessibility operations – see the Capital One blog) and others. These are paying customers that have an enterprise license,” said Founder and CEO, Navin Thandani.
Indeed, as Capitol One explains:
Capital One partnered with Evinced early, to guide their development with a particular focus on: helping developers release accessible code integrating multiple automated testing steps through the build and deployment lifecycle building products that can automatically scan for accessibility across a full web property (including through logins and internal repositories), and do this fast.
Capital One partnered with Evinced early, to guide their development with a particular focus on: helping developers release accessible code integrating multiple automated testing steps through the build and deployment lifecycle building products that can automatically scan for accessibility across a full web property (including through logins and internal repositories), and do this fast.
We’ve seen Evinced discover as much as 10x more critical accessibility issues than we were previously finding through automated testing alone. An even greater number of issues are discovered when a site is more interactive, including keyboard and screen reader usability issues.
Automated testing on a large enterprise scale can be an extremely complex and time consuming effort. Evinced is speedy and reliable, with 40x faster execution, enabling us to cut our processing time in some cases from 4-5 days down to less than 3 hours (and is being further optimized).
Glowing words, even if they are from an investor (technically Capital One Ventures, but still).
The company’s $17M series A was co-led by M12 BGV, and Capital One Ventures and included previous investors Engineering Capital.
As a sort of debut celebration present, Evinced is announcing its free tiers of service, including an iOS app accessibility debugger, which should be helpful to all the folks making apps who don’t know a thing about WCAG guidelines and ARIA roles. There’s also a free “community edition” site scanner that admins can sign up to be approved for, and a free trial for enterprises that want to give it a shot.
Source: https://techcrunch.com/2021/02/03/evinced-raises-17m-to-speed-up-accessibility-testing-for-the-web/
Amazon has started making deliveries to customers in Los Angeles using electric vans designed and built by Rivian.
The electric vans, which are part of Amazon’s 2040 climate pledge, won’t go into series production until the end of the year, according to an update Wednesday by the company. Amazon declined to reveal how many electric vans were in the test fleet.
The customer deliveries are part of continuous testing being conducted by Amazon and Rivian to measure performance as well as safety durability in various climates and geographies. Road tests first started more than four months ago. The current fleet of vehicles was built at Rivian’s headquarters in Plymouth, Michigan and can drive up to 150 miles on a single charge. Rivian engineers will continue to refine the vehicles for the start of production at its Normal, Illinois factory.
Image Credits: Amazon
In the meantime, these electric vehicles will continue to pop up on delivery routes in up to 15 additional cities in 2021. Eventually, Amazon plans to deploy at least 100,000 electric vans — the size of its order with Rivian — over the next several years.
Amazon and Rivian began testing vehicles four months prior to making customer deliveries, as part of the testing and development process. Amazon is also starting to modify its buildings to accommodate the new fleet of vehicles and has installed thousands of electric vehicle charging stations at its delivery stations across North America and Europe, the company said.
“We’re loving the enthusiasm from customers so far–from the photos we see online to the car fans who stop our drivers for a first-hand look at the vehicle,” Ross Rachey, director of Amazon’s global Fleet and products, said in a statement. “From what we’ve seen, this is one of the fastest modern commercial electrification programs, and we’re incredibly proud of that.”
The exterior of Rivian-built electric vans share some the same design features found in today’s gas-powered versions. There are a few more rounded edges and an overall sleeker look to the electric vans.
The real difference is in the electric architecture and the custom features that have been integrated into the vans, including, highway driving and traffic assist features; exterior cameras that can provide a 360-degree view for the driver via a digital display, a larger interior floor space in the cabin to help with drivers getting to and from the cabin compartment, surround tail lights for better braking visibility and three-level shelving and a bulkhead cargo compartment separating door. Amazon’s Alexa voice assistant is also an embedded feature.
Scratchpad is an early stage startup that wants to make it easier for sales people to get information into Salesforce by placing a notation layer on top of it. Today, it announced a $13 million Series A led by Craft Ventures with participation from Accel.
The company has now raised a total of $16.6 million including the $3.6 million seed round we covered in October. Co-founder and CEO Pouyan Salehi says that he wasn’t really looking to add capital, but the investors understood his vision and the money will help accelerate the product roadmap.
“To be honest, it actually wasn’t on our radar to raise again so soon after we raised what I consider a substantial seed. We had plenty of runway, but we started to see a lot of bottom-up user growth, this bottom-up motion just really started to take hold,” Salehi told me.
He says that lead investor David Sacks, who has built some successful startups himself, really got what they were trying to do, and the deal came together fairly easily. In fact, the company caught the attention of Craft because they were hearing about Scratchpad from their portfolio companies.
The bottoms up approach is certainly something we have seen with developer tools and with software for knowledge workers, but companies often take aim at sales through the sales manager, rather than trying directly to get salespeople to use a particular tool. This approach of getting the end users involved early allows them to gain traction with members of the sales team before approaching management about paid versions.
Traditionally, sales teams don’t like the tools that are thrust upon them. They are essentially databases and even with a visual interface, it doesn’t really match up with the way they work. Scratchpad gives them an interface like a spreadsheet or notes application that they are typically using to hack together a workflow, but with a direct connection to Salesforce.
What the paid tiers provide is a way to bring all this data together and get a bigger picture view of what’s happening on the sales team, and it helps ensure that people are using Salesforce because the data in Scratchpad links to the Salesforce database automatically.
The company has completed the initial work of building the individual salesperson’s workspace, but the next phase, and part of what this capital is going to fund, is building the team workspace and seeing how this data can flow from individuals to a team view to give management more insight into what their individual reps are doing. This includes notes, which usually don’t make it into Salesforce, but provide a lot of context about interactions with customers.
It’s resonating with thousands of users (although Salehi didn’t want to share an exact customer number just yet). Customers include Autodesk, Brex, Lacework, Snowflake and Twilio.
Sacks says that he liked the viral way the product has been spreading. “Once a rep starts using Scratchpad, two things tend to happen: it becomes a daily habit, and they share it with their teammates. This phenomena of viral spread is rare and indicates a very strong product-market fit,” he said in a statement.
Source: https://techcrunch.com/2021/02/03/scratchpad-snags-13m-series-a-to-simplify-salesforce-data-entry/
Training AIs is essential to today’s tech sector, but handling the amount of data needed to do so is intrinsically dangerous. DARPA hopes to change that by tapping the encryption experts at Duality to create a hardware-accelerated method of using large quantities of data without decrypting it — a $14.5 million contract.
Duality specializes in what’s called fully homomorphic encryption. Without descending into the technical details, the main issue with everyday encryption methods — though it’s also sort of the point of them — is that they render the encrypted data totally unreadable, essentially noise unless you have the key to reverse the process. Doing that is computationally expensive with large datasets, and of course once the data is in the clear, it’s vulnerable to hackers, abuse, and other dangers.
There are methods, however, of encrypting data such that it can be analyzed and manipulated without decrypting it, and one of those is fully homomorphic encryption. Unfortunately FHE is even more computationally intense than ordinary encryption, ruling it out for applications where gigabytes or terabytes of data are called for. There are other methods of accomplishing the same ends but no one would cry if FHE suddenly became ten times easier.
DARPA is as interested as anyone else in this field, though it has considerably deeper pockets than your garden variety encryption wonk. This contract is part of a broader effort called DPRIVE, or Data Protection in Virtual Environments, and the stated goal is to develop a special purpose chip — an ASIC pre-assigned the code name TREBUCHET — to accelerate FHE by, hopefully, an order of magnitude or more.
The Duality team will bring in experts from USC, NYU, CMU, SpiralGen, Drexel University, and TwoSix Labs. The company has been in the game for a long time and has actually worked with DARPA before, so this is not new territory for them.
“Duality team members have been supporting DARPA-funded innovation and application of FHE for over a decade. Some members of our team developed the first ever prototype HE hardware accelerators under the DARPA PROCEED program starting in 2010 and are lead developers for the PALISADE open source FHE library, first developed for the DARPA SAFEWARE program in 2015,” said Duality Labs director and principal investigator for the contract, David Bruce Cousins, in a press release.
As you can see, they’re not short on acronyms either.
It’s not totally clear what the timeline is on this, but considering the state of the technologies involved I wouldn’t expect results before at least two or three years from now.
If you’re at all interested in wine, chances are you’ve turned to Vivino at least a few times for recommendations. The app and the company behind it have been helping people enjoy better wine since 2010, and now the startup has raised $155 million with its Series D round – a sum over twice as large as all of its previous funding to date. Spurred by rapid growth that has seen its user base grow from 29 million in 2018, to 50 million currently, Vivino wants to use the large cash injection to significantly boost its core tech and personalized recommendation engine, while also expanding its presence in key growth markets globally.
Vivino is an interesting company for many reasons, but chief among them might be just how similar its vision today is to the one it started out with. Founder and CEO Heini Zachariassen told me in an interview that the app has been remarkably immune to the pivot – something as natural as breathing in the fast-flowing startup world.
“I can look at my slide, from when I pitched this 10 years ago,” he told me. “It says, ‘Hey, you scan a bottle of wine, then you can buy it.’ That just makes a lot of sense to anybody, so it really hasn’t changed much.”
“It’s been very, very difficult to build much – much harder to build than building that slide,” he joked. But it’s always been the same – we always knew that was going to be the model.”
That core value proposition is what leads to a lot of Vivino’s initial downloads and subsequent usage. The scenario is likely familiar: You’re sitting in a restaurant and browsing the wine menu, or staring at a crowded shelf in a wine store. For myself, I think I likely searched for something like ‘wine recommendation app’ and found Vivino via the App Store, installed it and was snapping photos of labels or menus within minutes. The recommendations provided somewhere to start, and since then the app has grown more personalized as I’ve provided input about my tastes.
Image Credits: Vivino
Vivino’s marketplace component means you can often buy the wines you find and enjoy directly from the app, via partnerships the company fosters and maintains with merchants large and small around the world. Zachariassen explained that they strive to maintain high standards when it comes to these partners, since the experience a user has with them is largely a reflection on Vivino itself because the app provides the means for the purchase.
Building more relationships with more merchants in more geographies is one part of their expansion goal for addressing their primary growth markets, but the company is also going to put a lot more capital behind improving and extending its recommendation engine. A lot of the building blocks are in place to make big improvements there, not least of which is the wine database that Vivino spent a decade building essentially from zero.
“Stage one of the hurdles we faced, even before we got commercial, was really building the data,” Zachariassen told me. “There is no aggregated data anywhere. So we’ve basically built this data totally from scratch. So it means taking a picture of bottle of wine, then having people just entering info every single day to fill it. We have 1.5 billion pictures of wine labels right now, so building that mass of data in a good and structured way really is 10 years of work.”
He adds that wine is a particularly long-tail marketplace, with highly individual tastes and very little indication in the company’s history that that’s likely to change in any significant way. Vivino’s marketplace approach, which is highly local on both the supply and the demand side, is particularly well-suited to addressing the sector’s needs, and Zachariassen believes Vivino has only really begun to scratch the surface on that thus far. I asked him why now was the right time to take on this sizeable round, given they’ve been very modest with prior funding amounts.
Vivino wine app in San Francisco. Photo Copyright Nader Khouri 2018.
“I think we we’ve reached sort of a critical mass,” he said. “We saw last year massive growth, and actually reaching […] like a quarter of a billion dollars in sales, and we’ve really seen that the unit economics are healthy for us. At the same time, unlike other marketplaces – you know, the order of things when you have a marketplace might be if you’re like Uber, is that you go into market, you spend money, do marketing, a lot of money to build up the demand, and then you build the supply on top. We’re a little bit different in the way that demand is already there, because we have 50 million users around the world. So we just follow our demand.”
“But the hardest thing about that is that we’re now a 200 person company that sells wine in 17 countries,” he continued. “Which means we’re relatively thin in all these markets. So so one of the big things here, is actually to go much deeper in each market and say, okay, we now know it works here, let’s put more resources in every single market.”
Zachariassen also added that the company spends very little on marketing to date, so it’s going to begin spending more on that to extend its organic growth. Finally, it wants to really build out product engineering, since he says that while users love the existing app, they really “want to do so much more with it.”
Vivino has worked to modernize a product category that has long relied on local expertise and individual storehouses of highly-specific information, with an approach that provides all the benefits of a connected and global marketplace, while retaining regional and particular appeal at the granola level of the individual user. Now, the company is read to tap the rest of the massive submerged demand it has identified, and this fresh fundings would help it do just that.
The $155 million series D round was led by Sweden’s Kinnevik, and also includes participation by Sprints Capital, GP BullHound, and existing investor Creandum which led its Series A. This brings the company’s total funding to $221 million to date.
