TechCrunch |
- HitPay is a one-stop solution for SMEs
- Zinc heads towards new $41M tech-for-good fund to back pre-team talent solving big problems
- Max Q: Mines and metals
- Partner sessions at TC Sessions: Climate offer knowledge and insight
- Perceptron: Risky teleoperation, Rocket League simulation, and zoologist multiplication
- For remittances, crypto is still a problem looking for a solution
- Strong Compute raises $7.8M seed round to speed up ML training pipelines
- SWVL plans to lay off 32% of its team two months after going public
- Mintlify taps AI to automatically generate documentation from code
- The Station: EV SPACs face new regulatory speed bump, more on Rivian’s reorg and VW weighs direct sales for Scout brand
- One AI raises $8M to curate business-specific NLP models
| HitPay is a one-stop solution for SMEs Posted: 30 May 2022 05:00 PM PDT HitPay has almost everything SMEs need to run their businesses. In addition to being an online payment gateway, it also offers tools like point-of-sale software with card readers, plugins, payment links and no-code online stores. The Y Combinator alum announced today that it has raised $15.75 million in Series A funding led by Tiger Global, with participation from returning investors Global Founders Capital and HOF Capital. It is currently used by over 10,000 merchants in Singapore and Malaysia, with plans to expand into more Southeast Asian markets, including Thailand, Indonesia and the Philippines. Co-founder and CEO Aditya Haripurkar told TechCrunch HitPay started in 2016 as an e-wallet, but then pivoted toward being a SME-facing platform in 2018 as a virtual POS product. As its team began to understand the needs of SMEs more, it started to develop the other tools on the platform. HitPay's Series A funding will be used for building a payments infrastructure from the ground up, with the intention of saving SMEs money and helping them expand their business. This will include business tools and payments infrastructure that includes all commonly used payment rails in each market, including bank transfers, cards, e-wallets and BNPL services. "SMEs have very specific requirements, so we wanted to build a one-stop no code platform," said Haripurkar. "That entails all our plugins, point of sale software, business software, online stores and recurring payments. We'll be focusing on building these free SaaS tools in addition to building up payment rails, which are focused currently on Singaporean and Malaysian merchants. But in each country we launch in, that will look very different, so we will look at local payment methods in every country. That's the biggest challenge for our team and where most of our investment and time is going as well." The first step HitPay will take as it expands into new countries is to get regulated in each market it operates in, to allow it to build payment infrastructure for SMEs from the ground up. Then it will integrate the most popular payment methods. For example, in Singapore, HitPay currently works with about 10 to 15 payment methods. HitPay's no-code platform allows SMEs to unify their online and offline payment stacks. It is typically used by medium-sized businesses, with annual revenue between $500,000 to $2 million. Most are in the retail segment, but Haripurkar expect that to evolve as well. |
| Zinc heads towards new $41M tech-for-good fund to back pre-team talent solving big problems Posted: 30 May 2022 04:01 PM PDT So-called "tech for good" accelerators addressing such worthy-sounding subjects as ESGs and SDGs have appeared in the last few years. Some observers have dismissed these efforts as scalable only put to a point. However, the evidence is mounting that they are increasingly attracting some of the world's best talent, because the world's best talent does actually want to solve some of the planet's biggest problems. And where the talent goes, the money and backing will follow. In Europe, Entrepreneur First (EF) and Antler have tried to scale their models as 'talent investors', while the Bethnal Green Ventures fund was even acquired and re-capitalized by its new owners. Clearly this approach is on something of a roll. Zinc is an accelerator which appeared back in 2017 when it was founded Ella Goldner, Paul Kirby and Saul Klein (LocalGlobe founder) and backed by its early investors including the London School of Economics. It went on to back over 220 diverse founders who built more than 60 ventures, such as Vira Health (menopause support), Tandem (transportation for workers), Pexxi (personalised contraception pills), and Untangle (grief support). As you can see, it is indeed possible to create businesses that address what, to some, might seem like intractable problems. Zinc has now hit the first £28m ($34m) close of a new fund, and is aiming for a final close of £33m ($41m), to invest in startups build commercial solutions to some of society's biggest challenges. Zinc will invest up to £250,000 in each of the companies created. Zinc 2 Fund will back talent which (similar to EF) is pre-team and pre-idea, to build these startups. The cohort draws in talent focused on four missions: mental health, the environment, improving the quality of later life, and helping people impacted by automation and globalisation. Zinc and the entrepreneurs share a conviction that each of these missions is a big opportunity for both social impact and commercial success. Goldner, co-founder of Zinc, said in a statement: "Rather than waiting for good companies to appear, Zinc helps individuals (before they have a business idea or a team) to build from scratch a new commercially-ambitious company to solve the social challenge that they are most passionate about." "Typically," she says "those individuals are 10 to 20 years into their career, but are frustrated that they are not having the social impact they want… Zinc brings these groups together to combine social impact and commercial skills." The individuals chosen by Zinc join a cohort of up to 70 people who all share the same mission and access a 12 month programme of support and investment. Each of our programmes has 100 Visiting Fellows and a network of partners. Givent the "Great Resignation" post pandemic, Zinc thinks it will attract those re-evaluating their careers. Paul Kirby, co-founder of Zinc, says "Our missions are a call to arms: 'Who wants to quit their jobs and spend the next decade or more solving this problem?'" Some of the examples of the founders who have built a venture with Zinc include Dr Rebecca Love, the co-founder of Vira Health which has raised $14m of VC funding, and Alex Shapland Howes of Tandem which has raised £2m. Other investors in Zinc’s new fund include Big Society Capital, Molten Ventures, Isomer, Dunhill Medical Trust, Atomico, Anthemis Group, Taavet Hinrikus from Taaven+Sten, Illka Paannan (Supercell), Basecamp, Sarah Wood and Stuart Roden. The founders Zinc backed in its most recent venture builder programme are over 50% women, 15% Black, with an average age of 38. |
| Posted: 30 May 2022 03:00 PM PDT Hello and welcome back to Max Q. There is SO much news this week, so let’s dive in. In this issue:
Don't forget to sign up to get the free newsletter version of Max Q delivered to your inbox. Astroforge closes $13M seed plus round for asteroid mining plansAlthough we've long understood that asteroids are not simply the rubble of the universe, but potentially profitable stores of precious minerals, humanity has never been able to unlock this value. Y Combinator startup Astroforge wants to succeed where other companies have failed, by becoming the first to mine an asteroid and bring the material back to Earth — and it's aiming to do so as early as the end of the decade. (Yes, that is not a typo — end of the decade!) To start, Astroforge will be conducting a tech demonstration mission sometime next year. The company’s already booked a spot on a SpaceX Falcon 9 rideshare mission, and it has also contracted a deal with OrbAstro to manufacture the satellite. But for now, the startup is staying mum on the actual details of the payload, and how they will solve the myriad technical challenges for which asteroid mining is so notorious. "We now need to build a world-class team to go after this, as it's a really hard problem to solve," co-founder Matt Gialich said. Later in the conversation, he added, "That's the fun part of startups, right? It's a big risk until you go do it."  Near-Earth asteroid, computer artwork. Image Credits: Science Photo Library – ANDRZEJ WOJCICKI / Getty Images Starliner returns to Earth, completing successful test flightWelcome back, Starliner! The spacecraft touched down in New Mexico on Wednesday, successfully concluding a six-day mission and the craft’s first successful test flight. As TC’s Devin Coldewey writes, even though not everything went exactly as planned, “this success may establish Boeing as a much-needed second provider of commercial ISS launch capabilities.” During a post-launch briefing, NASA commercial crew program manager Steve Stich called the landing “picture perfect.” Next up is a Crew Flight Test (CFT), which will carry astronauts and, for that reason, will be much higher stakes. The date of that launch will likely not be set for several months.  Image Credits: Boeing / NASA / YouTube More news from TC
…and beyond
Photo of the week I loved this photo, tweeted out by Relativity Space, of Terran 1’s second stage crossing state lines. If all goes to plan, Terran 1 will make its first orbital launch attempt by the end of this year. Image Credits: Relativity Space (opens in a new window) Max Q is brought to you by me, Aria Alamalhodaei. If you enjoy reading Max Q, consider forwarding it to a friend. |
| Partner sessions at TC Sessions: Climate offer knowledge and insight Posted: 30 May 2022 02:30 PM PDT We're just about two weeks away from our first foray into climate tech at TC Sessions: Climate & The Extreme Tech Challenge 2022 Global Finals on June 14 in Berkeley, California — with an online day to follow on June 16. It's going to be an epic day all around for many reasons — did you know that Bill Gates is one of the featured speakers? Don't miss this opportunity to hear from and engage with the new wave of climate-tech entrepreneurs, early-stage founders, CEOs, scientists, researchers, engineers and the VCs who fund them. You know what else you can't afford to miss? Our 2-for-1 pass Memorial Day sale — it ends tonight at 11:59 pm (PT). Buy your pass now and save! Pro Tip: Yes, TechCrunch editors will interview the leading voices in the fight against climate-change (check out the event agenda), but we’d be doing you a disservice if we didn't remind you about our partner breakout sessions. These expert-led, topic-specific partner sessions give you time to lean in, get more answers, discover new opportunities and connect with companies that support early-stage climate-tech startups. Take a look at this impressive group of partners and what they'll discuss. They're ready with knowledge and resources to help you build a successful startup. Powering the Future Through Transformative TechThis panel jumps into the breakthrough tech innovations that are transforming industries to build a radically better world. How can business, government, philanthropy, and the startup community come together to create a better tomorrow? Hear from these seasoned investors and industry veterans about how technology can not only shape the future, but also where the biggest opportunities lie. Sponsored by XTC. Speakers: Reducing your cloud computing climate impactMaking the choice to deploy to the cloud is clearly the better choice for the climate, but you can further reduce your emissions by taking a couple of key steps. You are invited to attend this session to learn more about how the tech community is helping to mitigate climate change and a simple strategy to reduce your carbon footprint in the cloud. Sponsored by Platform.sh Speakers: TC Sessions: Climate 2022 takes place on June 14 in Berkeley, California (with an online day June 16). Be sure to soak up the knowledge waiting for you in our partner breakout sessions. And don't wait another second to buy your pass — the Memorial Day 2-for-1 sale ends tonight at 11:59 pm (PT). Go forth and save! Is your company interested in sponsoring or exhibiting at TC Sessions Climate 2022? Contact our sponsorship sales team by filling out this form.
|
| Perceptron: Risky teleoperation, Rocket League simulation, and zoologist multiplication Posted: 30 May 2022 10:48 AM PDT Research in the field of machine learning and AI, now a key technology in practically every industry and company, is far too voluminous for anyone to read it all. This column, Perceptron (previously Deep Science), aims to collect some of the most relevant recent discoveries and papers — particularly in, but not limited to, artificial intelligence — and explain why they matter. This week in AI, researchers discovered a method that could allow adversaries to track the movements of remotely-controlled robots even when the robots’ communications are encrypted end-to-end. The coauthors, who hail from the University of Strathclyde in Glasgow, said that their study shows adopting the best cybersecurity practices isn’t enough to stop attacks on autonomous systems. Remote control, or teleoperation, promises to enable operators to guide one or several robots from afar in a range of environments. Startups including Pollen Robotics, Beam, and Tortoise have demonstrated the usefulness of teleoperated robots in grocery stores, hospitals, and offices. Other companies develop remotely-controlled robots for tasks like bomb disposal or surveying sites with heavy radiation. But the new research shows that teleoperation, even when supposedly “secure,” is risky in its susceptibility to surveillance. The Strathclyde coauthors describe in a paper using a neural network to infer information about what operations a remotely-controlled robot is carrying out. After collecting samples of TLS-protected traffic between the robot and controller and conducting an analysis, they found that the neural network could identify movements about 60% of the time and also reconstruct “warehousing workflows” (e.g., picking up packages) with “high accuracy.”  Image Credits: Shah et al. Alarming in a less immediate way is a new study from researchers at Google and the University of Michigan that explored peoples’ relationships with AI-powered systems in countries with weak legislation and “nationwide optimism” for AI. The work surveyed India-based, “financially stressed” users of instant loan platforms that target borrowers with credit determined by risk-modeling AI. According to the coauthors, the users experienced feelings of indebtedness for the “boon” of instant loans and an obligation to accept harsh terms, overshare sensitive data, and pay high fees. The researchers argue that the findings illustrate the need for greater “algorithmic accountability,” particularly where it concerns AI in financial services. “We argue that accountability is shaped by platform-user power relations, and urge caution to policymakers in adopting a purely technical approach to fostering algorithmic accountability,” they wrote. “Instead, we call for situated interventions that enhance agency of users, enable meaningful transparency, reconfigure designer-user relations, and prompt a critical reflection in practitioners towards wider accountability.” In less dour research, a team of scientists at TU Dortmund University, Rhine-Waal University, and LIACS Universiteit Leiden in the Netherlands developed an algorithm that they claim can “solve” the game Rocket League. Motivated to find a less computationally-intensive way to create game-playing AI, the team leveraged what they call a “sim-to-sim” transfer technique, which trained the AI system to perform in-game tasks like goalkeeping and striking within a stripped-down, simplified version of Rocket League. (Rocket League basically resembles indoor soccer, except with cars instead of human players in teams of three.)  Image Credits: Pleines et al. It wasn’t perfect, but the researchers’ Rocket League-playing system, managed to save nearly all shots fired its way when goalkeeping. When on the offensive, the system successfully scored 75% of shots — a respectable record. Simulators for human movements are also advancing at pace. Meta’s work on tracking and simulating human limbs has obvious applications in its AR and VR products, but it could also be used more broadly in robotics and embodied AI. Research that came out this week got a tip of the cap from none other than Mark Zuckerberg.  Simulated skeleton and muscle groups in Myosuite. MyoSuite simulates muscles and skeletons in 3D as they interact with objects and themselves — this is important for agents to learn how to properly hold and manipulate things without crushing or dropping them, and also in a virtual world provides realistic grips and interactions. It supposedly runs thousands of times faster on certain tasks, which lets simulated learning processes happen much quicker. “We’re going to open source these models so researchers can use them to advance the field further,” Zuck says. And they did! Lots of these simulations are agent- or object-based, but this project from MIT looks at simulating an overall system of independent agents: self-driving cars. The idea is that if you have a good amount of cars on the road, you can have them work together not just to avoid collisions, but to prevent idling and unnecessary stops at lights.  If you look closely, only the front cars ever really stop. As you can see in the animation above, a set of autonomous vehicles communicating using v2v protocols can basically prevent all but the very front cars from stopping at all by progressively slowing down behind one another, but not so much that they actually come to a halt. This sort of hypermiling behavior may seem like it doesn’t save much gas or battery, but when you scale it up to thousands or millions of cars it does make a difference — and it might be a more comfortable ride, too. Good luck getting everyone to approach the intersection perfectly spaced like that, though. Switzerland is taking a good, long look at itself — using 3D scanning tech. The country is making a huge map using UAVs equipped with lidar and other tools, but there’s a catch: the movement of the drone (deliberate and accidental) introduces error into the point map that needs to be manually corrected. Not a problem if you’re just scanning a single building, but an entire country? Fortunately, a team out of EPFL is integrating an ML model directly into the lidar capture stack that can determine when an object has been scanned multiple times from different angles and use that info to line up the point map into a single cohesive mesh. This news article isn’t particularly illuminating, but the paper accompanying it goes into more detail. An example of the resulting map is visible in the video above. Lastly, in unexpected but highly pleasant AI news, a team from the University of Zurich has designed an algorithm for tracking animal behavior so zoologists don’t have to scrub through weeks of footage to find the two examples of courting dances. It’s a collaboration with the Zurich Zoo, which makes sense when you consider the following: “Our method can recognize even subtle or rare behavioral changes in research animals, such as signs of stress, anxiety or discomfort,” said lab head Mehmet Fatih Yanik. So the tool could be used both for learning and tracking behaviors in captivity, for the well-being of captive animals in zoos, and for other forms of animal studies as well. They could use fewer subject animals and get more information in a shorter time, with less work by grad students poring over video files late into the night. Sounds like a win-win-win-win situation to me.  Image Credits: Ella Marushenko / ETH Zurich Also, love the illustration. |
| For remittances, crypto is still a problem looking for a solution Posted: 30 May 2022 08:00 AM PDT It is no secret that Andreessen Horowitz is bullish about crypto: Not only does the firm boast that it started to invest in the space a decade ago, but it also debuted a $4.5 billion web3 fund last week. To understand a16z’s bullishness despite what others have described as a “crypto winter,” its 2022 State of Crypto Report is a good start. Per its disclaimers, the document is not directed to any investors or potential investors — yada, yada, yada. But it does read like an argument for crypto, DeFi, NFTs and all things web3. The problem, in my view, is that the report’s authors, all of whom are part of a16z’s team, are overstating the current opportunity for crypto. By doing so, they are making it sound bigger than it is — and it may take years to get to that point. TechCrunch+ is having a Memorial Day sale. You can save 50% on annual subscriptions for a limited time. That the report takes an optimistic view of crypto is understandable. After all, if you are about to deploy billions in funding into a market, and you are not even alone, the TAM needs to be up to par. But the report is also meant to be an overview of trends, which is why it seems questionable to allude to opportunities that aren’t real yet. The point that bothered me the most has to do with remittances — money sent cross-border by individuals, typically from a richer country to a poorer one. The World Bank expects that such annual inflows will reach $630 billion in 2022. And yes, there are inefficiencies and fees along the way. For the authors of the a16z report, that’s more than enough to list remittances as an argument for DeFi. But are remittance payments and money transfers really ripe for crypto disruption? And is DeFi really the right solution to help what the report describes as the “huge part of the world [ … ] underserved by existing financial institutions”? That’s definitely not what I am hearing from the ground — as also confirmed by two founders I reached out to Tomás Bercovich from Global66 and Ryan Newton from Paisa. Thanks, but no thanksJust earlier this month, I sat in the audience of the Tech.eu Summit as Wise CEO Kristo Käärmann was interviewed on stage. “Currently, Wise doesn’t accept cryptocurrencies. Do you think,” Bloomberg’s Ivan Levingston asked him, “that this might change at some point?” This is a recurring question for the fintech company, so Käärmann made sure not to sound dismissive. “I am very excited about the technology,” he said, while also adding that “there are interesting experiments going on all around the world.” But the gist of his answer was still a nail in the crypto coffin. “We are just looking for a use case,” he said. “We’re looking for the problem that we can solve with it.” |
| Strong Compute raises $7.8M seed round to speed up ML training pipelines Posted: 30 May 2022 07:00 AM PDT Strong Compute, a Sydney, Australia-based startup that helps developers remove the bottlenecks in their machine learning training pipelines, today announced that it has raised a $7.8 million seed round. The round includes a total of 30 funds and angels, including the likes of Sequoia Capital India, Blackbird, Folklore and Skip Capital, as well as Y Combinator, Starburst Ventures and founders and engineers from companies like Cruise, Waymo, Open AI, SpaceX and Virgin Galactic. The company, which was part of Y Combinator’s Winter ’22 batch, promises that its optimizations can speed up the training process by 10x to 1000x, depending on the model, pipeline and framework. As Strong Compute founder Ben Sand, who previously also co-founded AR company Meta, told me, the team has recently made some breakthroughs where it was able to take Nvidia’s reference implementation, which its customer LayerJot used, to run 20 times faster.  Image Credits: Strong Compute “That was a big win,” Sand said. “It really gave us the sense that there is nothing that can’t be improved.” He didn’t quite want to reveal all of the details of how the team’s optimizations worked, but he noted that the company is now hiring mathematicians and is building tools that give it a more detailed view of how their user’s code interacts with the CPUs and GPUs at a much deeper level than was previously possible. As Sand stressed, the current focus for the company is to start automating a lot of the current work to optimize the training process — and that’s something the company can now tackle, thanks to this funding round. “Our goal now is to have serious development partners in self-driving, medical and aerial, in order to be looking at what is actually going to generalize really well,” he explained. “We’ve now got the resources to have an R&D team that doesn’t have to deliver something in a two-week sprint but that can actually look at what’s some real core tech that could take a year to actually get a win out of but that can really help with that automated analysis of the problem.” The company currently has six full-time engineers but Sands plans to double that over the next few months. In part, that’s also because the company is now getting inbound interest from large companies that often spend $50 million or more on their compute resources (and Sands noted that the market is basically bi-modal, with customers either spending less than $1 million or $10 to $100 million, with only a few players in the middle).  Image Credits: Strong compute Every company that is trying to build ML models, though, suffers from the same problem: training models and running experiments to improve them still take a lot of time. That means the well-paid data scientists working on these problems spend a lot of time in a holding pattern, waiting for results to come in. “Strong Compute is solving the basketball court problem,” said SteadyMD CFO Nikhil Abraham. “Long training times had our best devs shooting hoops all day, waiting on machines.” And while some of that inbound interest is coming from the financial industry and companies that want to optimize their natural language processing models, Strong Compute’s focus remains on computer vision for the time being. “We’ve only just scratched the surface of what machine learning and AI can do.” said Folklore partner Tanisha Banaszcyk. “We love working with founders who have long-range ambition and visions that will endure across generations. Having invested in autonomous driving, we know how important speed to market is – and see the impact Strong Compute can have on this market with its purpose-built platform, deep understanding of the $500B market and world-class team.” |
| SWVL plans to lay off 32% of its team two months after going public Posted: 30 May 2022 06:05 AM PDT Egypt-born and Dubai-headquartered mobility startup SWVL is planning to lay off 32% of its workforce, it said in a statement today. The company’s LinkedIn profile shows it has over 1,330 employees. Letting go of over 30% of its workforce means that around 400 people will lose their job at the mobility company. Tech companies, private and public, have faced a reckoning in the past few months with their valuations taking a beating. The effect of an economic downturn has also affected their finances leading them to cut costs; the top of the list is letting go of employees. This downsizing from the Dubai-based startup adds to the long list of global cross-stage layoffs in what has been a rough month for tech employees. Over 15,000 tech workers have lost their jobs in the U.S. alone according to reports. Companies such as Klarna, Getir, Gorillas and Bolt have dismissed portions of their workforce while the likes of Snap, Twitter and Instacart have slowed down hiring entirely. It’s been a very busy 18 months for SWVL leading up to this news. This March, the company went public via a SPAC merger with U.S. women-led blank check company Queen's Gambit Growth Capital. It listed at $10 per share and targeted a $1.5 billion valuation but has traded between $4 and $8 for the most part. Its current valuation hovers around $500-600 million. The layoffs are coming just a month after SWVL acquired U.K.-based mass transit group Zeelo for $100 million according to sources. It’s one of five acquisitions SWVL has made within the past year; others include Germany's door2door, Turkey's Voltlines (for ~$40 million), Spain's Shotl and Argentina's Viapool. SWVL said that though these acquisitions have contributed to its overall growth, it will need to make reductions on roles automated by investments in its engineering and product and support functions teams. “The planned layoffs will impact teams responsible for functions that have been automated following investment in engineering, product and support functions,” SWVL said in a statement. SWVL said it plans to attain profitability next year. Dismissing hundreds of employees is one way to get there. Others include developing its proprietary technology stack and growing its three models — where it makes $5 million in MRR — across existing and new markets, it said in a statement. SWVL is present in 13 markets globally: the UAE, Egypt, Kenya, Germany, Spain, Italy, Switzerland, Turkey, Japan, Argentina, Saudi Arabia, Jordan, and Pakistan. According to a source, the majority of the layoffs will come from the company's Dubai and Pakistan offices. Whether SWVL will continue its expansion into new markets such as Colombia, Mexico and South Africa, and the U.S. — announced during it’s SPAC merger — is uncertain. “Swvl plans to provide monetary, non-monetary and job placement support to help transition certain of its employees to new roles,” the company said in a statement on how it plans to support affected employees. “As a result of the portfolio optimization program, Swvl's management currently expects that the company will be cash-flow positive in 2023.” CEO Mostafa Kandil sent out a letter to his employees addressing the layoffs. Here’s a part of it:
Update: In an email to TechCrunch, CFO Youssef Salem said Swvl is not shutting down operations in any country whether from its existing footprint or planned expansions but rather optimizing its network and headcount in each country. |
| Mintlify taps AI to automatically generate documentation from code Posted: 30 May 2022 05:00 AM PDT Mintlify, a startup developing software to automate software documentation tasks, today announced that it raised $2.8 million in a seed round led by by Bain Capital Ventures with participation from TwentyTwo Ventures and Quinn Slack, Sourcegraph's co-founder. CEO Han Wang says that the proceeds will be put toward product development and doubling Mintlify’s core, three-person team by the end of the year. Ithaca, New York-based Mintlify was co-founded in 2021 by Han Wang and Hahnbee Lee — both software engineers by trade. Wang previously co-launched Foodful, an startup that developed a cloud-based monitoring system for cows, and Pe•ple, an online customer community platform that was acquired by Tribe in early 2021. Lee was a co-founder at Pe•ple before briefly joining Duolingo as an engineer. Wang said the idea for Mintlify came from his and Lee’s experiences in software development, which involved working with documentation that wasn’t always complete or of the highest quality. Their observations agree with a 2017 GitHub survey, which found that 93% of developers consider incomplete or outdated documentation to be a pervasive problem. “We've worked as software engineers at companies in all stages ranging from startups to big tech and found that they all suffer from bad documentation, if it even existed at all,” Wang told TechCrunch in an email interview. “Documentation is the lifeline for junior engineers and those jumping into new codebases. It helps senior devs save time from explaining their code to others in the future. For public-facing and open-source products, documentation has a direct impact on user adoption.”  Image Credits: Mintlify Mintlify aims to address the challenges around documentation with automation, specifically auto-generating documentation. The company’s platform reads code and creates docs to explain it, leveraging technologies including natural language processing and web scraping. Wang declined to reveal more about Mintlify’s technical underpinnings, but generating documentation from code is well within the realm of possibility with today’s AI techniques. That’s evidenced by the fact that Mintlify has several competitors taking similar approaches, including Documatic, whose AI system automatically generates changelogs and explanations from code in addition to documentation. Wang asserts that Mintlify provides “significantly” higher-quality results than its rivals, and — unlike some — doesn’t force developers to host documentation on a cloud service. “Mintlify's mission is to solve documentation rot by developing continuous documentation into a standard practice for software teams,” Wang added. “[W]hen engineering managers are actively seeking solutions for better documentation practices, … that's when we step in.” Beyond generating documentation, Mintlify routinely scans for “stale” documentation and detects how users engage with the documentation to improve its readability. (Wang emphasized that the platform doesn’t store code and encrypts all user data at rest and in transit.) Mintlify, which is free for individual developers, also integrates with existing systems including Slack, Dropbox, and GitHub for automating task management and development workflows. According to Wang, adoption of Mintlify’s free plan has been growing 20% every week since its January launch. With the user base now eclipsing 6,000 active accounts, the company plans to shift its focus to a premium offering oriented toward enterprise customers. “The pandemic standardized a decentralized and asynchronous work environment. This made high-quality documentation critical to achieving efficient communication, onboarding, and product development,” Wang said. “Our expansion into workflow automations is addressing the challenge by targeting the engineering managers with our existing fanbase serving as champions.” |
| Posted: 30 May 2022 04:00 AM PDT Welcome back to The Station, your weekly guide to everything going on in the world of transportation. Many readers of this weekly newsletter are likely enjoying a three-day weekend thanks to the Memorial Day holiday. So, this week I will keep it a wee bit shorter. Before we get started, check out the latest transportation Q&A, this time with Convoy co-founder and CEO Dan Lewis. He predicts digital freight will go mainstream within the year. Why? Lewis said: “The industry is going to contract differently. Brokers are going to think about digital capacity and how they get access to that; they're going to be comfortable using a platform like Convoy for that.” As always, you can email me at kirsten.korosec@techcrunch.com to share thoughts, criticisms, opinions, or tips. You also can send a direct message to @kirstenkorosec or @rebeccabellan. Micromobbin’
Maybe scooter ADAS isn't such a bad idea after all. A new report from the UK Department for Transportation revealed that the number of pedestrians injured after being hit by e-scooters was nearly four times higher in 2021, at 223 people, than in 2020 when 57 people were injured. Seeing as how it's still illegal in the UK for people to ride private e-scooters, most of these injuries occurred with vehicles from shared micromobility schemes. The report doesn't say where these accidents occurred – were scooter riders on the sidewalk, and if so, are there protected bike lanes on the street where the accidents occurred? PeopleForBikes, in partnership with Wend Collective, released a report celebrating the milestones of certain U.S. cities that have systematically increased bike mobility by following the organization's grassroots effort strategy. Here are some of the highlights: In other news…A BCG survey predicts subscription will be the fastest-growing area of micromobility this decade. Self-combusting micromobility vehicles are getting more abundant. Serial 1 has released its new BASH/MTN e-bike, a single-speed, small-batch production bike where minimalism meets off-road adventure. Squad Mobility launched pre-orders for its Solar City Car, a two-seater compact city car that charges itself, in part, on solar energy. The car will be available in 2023 from €6,250. See ya next week! — Rebecca Bellan A little bird
Perhaps you saw coverage of a leaked internal email, first viewed by Bloomberg, from Rivian founder and CEO RJ Scaringe that detailed changes among some of its higher profile leaders at the company. Initial reports, including ours, focused on Rivian hiring Frank Klein as its new chief operating officer. Klein will join the company June 1 to oversee the automaker’s production, manufacturing engineering and supply chain. The automaker also said its head of manufacturing engineering, Charly Mwangi, is leaving the company. I finally got my hands on the complete internal email and there are a few other changes worth noting. Rivian is dividing its commercial business with one group focused on its B2B vertical and the other on its B2C vertical. Chief Growth Officer Jiten Behl will focus on the B2B business, which includes commercial vans, Rivian’s fleet management software FleetOS, charging, and importantly, its relationship with Amazon. (And that makes sense considering Behl is largely responsible for Rivian’s relationship with Amazon.) Behl, who will report directly to Scaringe, will also oversee strategy and corporate development, which includes strategic partnerships, business development activities and launching in international markets. Rivian has launched a job search for a Chief Commercial Officer to lead its B2C business. Rivian is also changing up its marketing, brands and comms team a bit. The brand team led by Forest Young and the communications team, led by Nick Mulholland, will now report directly to Scaringe. The marketing team, led by Lindsey Pearl, will be join the communications team. Then over in manufacturing and operations is the new COO I noted above. Under Klein, will be several executives, including Tim Fallon, who heads up manufacturing operations, Michael Smith, who is VP of quality, and Steve Gawronski, who is VP of supply chain. The manufacturing engineering spot, which Mwangi previously held until resigning, will also move under the COO. Finally, Jimmy Knauf, the executive VP of facilities, and Rivian’s entire facilities team will join the CFO organization and report directly to Claire McDonough. Deal of the week
Just a roundup this week, folks! American Axle & Manufacturing Holdings is exploring a potential sale, Bloomberg reported, citing unnamed people familiar with the matter. The Detroit-based auto-parts supplier has already hired an adviser to help with the process and a leveraged buyout through a private equity firm is a likely option. Battery Ventures, the U.S. private equity firm, made a A$657.6 million ($466.76 million) preliminary buyout offer for Australia-based automotive software provider Infomedia, Reuters reported. The A$1.75 in cash per Infomedia share bid is 5 Australian cents more than the rival proposal from a consortium led by U.S. private equity firm TA Associates. Booster, the on-demand direct-to-fuels delivery startup, raised more than $125 million in Series D funding led by Rose Park Advisors with participation from new investors Chaac Ventures, Equinor Ventures, Mitsubishi Corporation and Thayer Ventures. Renewable Energy Group, a producer of low-carbon renewable fuels, also increased its strategic investment. Existing investors Cercano Management, Conversion Capital, Enterprise Holding Ventures, Invus Opportunities, Madrona Venture Group, Maveron Ventures, Perot Jain LP and Version One Ventures also participated. BMW i Ventures led an undisclosed investment in the Series A-1 financing round of Mangrove Lithium. BMW i Ventures was joined by existing investor Breakthrough Energy Ventures in the round. Caribou, the auto financing and insurance platform, raised $115 million in Series C funding led by Goldman Sachs. The funding has pushed the fintech company’s valuation past the $1 billion mark. Other investors in the round included Innovius Capital, Harmonic, Accomplice, CMFG Ventures, Curql Collective, Firebolt Ventures, Gaingels, Moderne Ventures and Motley Fool Ventures. Finn, the car subscription startup, raised $110 million in a Series B round led by Korelya Capital along with Keen Venture Partners, Climb Ventures, Greentrail Capital and Waterfall Asset Management. Existing investors such as White Star Capital, HV Capital, Heartcore Capital, UVC Partners and Picus Capital also participated in the round. The funds will be used to help Finn expand in the U.S. and Europe and reach 30,000 subscriptions by the end of the year. Johnson Matthey agreed to most of its battery material unit to Australia’s EV Metals Group for £50 million ($63.15 million), reported the Financial Times. Meanwhile, Johnson Matthey sold its battery materials unit in Canada to Nano One for C$10.25 million ($8.05). Meritor announced last week (I missed this one) that it agreed to acquire Siemens’s 200-person Commercial Vehicles business for approximately €190 million ($203.9 million) in cash. Motive, the fleet management software startup that formerly known as KeepTruckin, raised $150 million in series F funding round co-led by Insight Partners and Kleiner Perkins. Motive’s post-funding valuation is now $2.85 billion, according to the company. Natrion, a solid-state battery technology startup founded in 2019, raised $2 million in seed funding ed by Chicago-based TechNexus Venture Collaborative with participation from Tamarack Global, Mark Cuban, Illinois Ventures, and other private investors. Nowports, an automated digital freight forwarder in Latin America, raised $150 million in a Series C funding round led by SoftBank Latin America Fund. The raise, which pushes its post-funding valuation to $1.1 billion, comes just six months after the Monterey, Mexico-based startup announced it had secured $60 million in a Series B funding round led by Tiger Global Management. Roetz, a Dutch e-bike maker, has raised $1.9 million from from Veth Investments and entrepreneurs Pim Claassen and Alexander van Citters. Sylndr, the Cairo-based online used car startu, raised a pre-seed round of $12.6 million — the largest of its kind in MENA and sub-Saharan Africa. Saudi-based early-stage venture capital firm RAED Ventures led the round. VC firms Algebra Ventures, Nuwa Capital, 1984 Ventures and Global Founders Capital participated, with several regional and global angel investors taking part as well. United Gear & Assembly, a maker of high-precision gears for EVs and other industries, has agreed to merge with special purpose acquisition company Aesther Healthcare Acquisition Corp. United Gear, a subsidiary of United Star Holdings, designs and makes high-precision gears for multiple large end markets and its customers include Allison Transmission, Caterpillar, GM, Ingersol Rand, John Deere, Lucid Group, Toyota and Volvo. Notable reads and other tidbits |
| One AI raises $8M to curate business-specific NLP models Posted: 30 May 2022 03:00 AM PDT Whether to power translation to document summarization, enterprises are increasing their investments in natural language processing (NLP) technologies. According to a 2021 survey from John Snow Labs and Gradient Flow, 60% of tech leaders indicated that their NLP budgets grew by at least 10% compared to 2020, while a third said that spending climbed by more than 30%. It’s a fiercely competitive market. Beyond well-resourced startups like OpenAI, Cohere, AI21 Labs, and Hugging Face and tech giants including Google, Microsoft, and Amazon, there’s a new crop of vendors building NLP services on top of open source AI models. But Yochai Levi isn’t discouraged. He’s one of the co-founders of One AI, an NLP platform that today emerged from stealth with $8 million led by Ariel Maislos, Tech Aviv, Sentinel One CEO Tomer Wiengarten, and other unnamed venture firms and angel investors. “While the market is growing fast, advanced NLP is still used mainly by expert researchers, big tech, and governments,” Levi told TechCrunch via email. “We believe that the technology is nearing its maturity point, and after building NLP from scratch several times in the past, we decided it was time to productize it and make it available for every developer.”  Image Credits: One AI Levi lays out what he believes are the major challenges plaguing NLP development. It’s often difficult to curate open source models, he argues, because they have to be matched both to the right domain and task. For example, a text-generating model trained to classify medical records would be a poor fit for an app designed to create advertisements. Moreover, models need to be constantly retrained with new data — lest they become “stale.” Case in point, OpenAI’s GPT-3 responds to the question “Who’s the president of the U.S.?” with the answer “Donald Trump” because it was trained on data from before the 2020 election. Levi believes the solution is a package of NLP models trained for particular business use cases — in other words, One AI’s product. He teamed up with CEO Amit Ben, CPO Aviv Dror, and CSO Asi Sheffer in 2021 to pursue the idea. Ben previously was the head of AI at LogMeIn after the company acquired his second startup, Nanorep, an AI and chatbot vendor. Dror helped to co-found Nanorep and served as a platform product manager at Wix. Sheffer, a former data scientist at Nanorep, was the principal data scientist at LogMeIn. As for Levi, he was the VP of online marketing at LivePerson and the head of marketing at WeWork. One AI offers a set of models that can be mixed and matched in a pipeline to process text via a single API call. Each model is selected and trained for its applicability to the enterprise, Levi said, and automatically matched by the platform to a customer’s task and domain (e.g., conversation summarization, sales insights, topic detection, and proofreading). One AI’s models can also be combined with open source and proprietary models to extend One AI’s capabilities. The platform’s API accepts text, voice, and video inputs of various formats. With One AI’s language studio, users can experiment with the APIs and generate calls to use in code. “With the maturation of Language AI technologies, it is finally time for machines to start adapting to us,” Ben told TechCrunch via email. "The adoption of language comprehension tools by the broader developer community is the way to get there.”  Image Credits: One AI One AI prices its NLP service across several tiers, including a free tier that includes processing for up to one million words a month. The stackable “growth tier” adds 100,000 words for $1. The hurdle One AI will have to overcome is convincing customers that its services are more attractive than what’s already out there. In April, OpenAI said that tens of thousands of developers were using GPT-3 via its API to generate words for over 300 apps. But, with Fortune Business Insights pegging the NLP market at $16.53 billion in 2020, it could be argued that there’s a large enough slice of the pie for newcomers. Added TehAviv founder and managing partner Yaron Samid: “Language is the most valuable untapped resource, beyond the reach of most products and companies. This is true to most industries and domains and results in negative outcomes that include everything from lost sales, to lower levels of user engagement and loyalty, to reputation damage. Unleashing Language AI allows us to harness the power of this unstructured data and turn it into useful information and insights.” One AI says that a portion of the seed round proceeds will be put toward expanding its 22-person team, which includes 10 NLP data scientists. |
| You are subscribed to email updates from TechCrunch. To stop receiving these emails, you may unsubscribe now. | Email delivery powered by Google |
| Google, 1600 Amphitheatre Parkway, Mountain View, CA 94043, United States | |
No comments:
Post a Comment