Tech News

As companies build more complex machine learning models, the cost of training and running these models becomes a real issue. AWS has created a series of custom instances to help bring down the cost, and today it introduced a preview of an all-new Inf2 instance for EC2 designed to process data from larger workloads more efficiently.

AWS CEO Adam Selipsky made the announcement today at AWS re:Invent in Las Vegas

As Selipsky explained, “Inf1 is great for small-to-medium complexity models, but for larger models, customers have often relied on more powerful instances because they don’t actually have the optimal resource configuration for their inference workloads,” he told the AWS re:Invent audience.

They did this because up until now, there simply wasn’t another solution available to help bring down the cost and complexity of processing these larger workloads.

“You want to choose the solution that is the best fit for your specific needs, which is why today I’m excited to announce a preview of the Inf2 instance powered by our new inferential two chip,” he said.

For folks who need that extra power, Inf2 provides it. “Customers can deploy a 175 billion parameter model for inference on a single instrument with four times higher throughput and 1/10 the latency of Inf1 instances,” he said.

The new instances are available in preview starting today.

Amazon announces preview of new Inf2 instances designed for larger models by Ron Miller originally published on TechCrunch

Plenty of companies that launched during downturns went on to be phenomenally successful.

During the Great Depression, Stanford grads David Packard and William Hewlett famously set up shop in a Palo Alto garage. Microsoft was founded as the U.S. was recovering from a years-long oil embargo that hobbled the economy. Slack, Airbnb, Uber and Square all rose from the ashes of the Great Recession.

As of September 2022, investors have amassed almost $300 billion in dry powder, and VC funds are still raising money by the boxcar. That’s because even during recessionary times, VC funds tend to outperform public markets.

Which explains why I’ve never heard an investor say it’s a bad time to launch a startup. But ask a few entrepreneurs, and you may get a different story.

Full TechCrunch+ articles are only available to members.
Use discount code TCPLUSROUNDUP to save 20% off a one- or two-year subscription.

According to a pre-seed report by DocSend, founders took an average of 52 meetings with investors in 2022 compared to 39 last year. At the same time, they are submitting 30% more pitch decks, but VC engagement has fallen 23%.

In Q4 2022, it takes more time to raise less money.

“Founders may be discouraged in this environment, but they need to remember that they have ‘currency,’ too,” said Russ Heddleston, co-founder and former CEO of DocSend at Dropbox.

Because investors spend less time reviewing pitches, concise, data-driven storytelling is more important than ever. DocSend’s report recommends using no more than 50 words per slide.

The sections of the deck that address purpose, product and business model are the meat in the sandwich, so founders should spend the most time polishing those points.

“Investors spent the third-highest amount of time reviewing the company purpose slide in pre-seed pitch decks, behind only the business model and product slides,” said Heddleston.

The idea that there’s a “good” time to launch a startup is just a bedtime story investors tell founders, and I regret any role I had in promoting it. Starting a company is an uphill slog on an uncertain path, and it’s not for everyone.

But if that’s your path, don’t let anyone talk you out of it.

Thanks for reading.

Walter Thompson
Editorial Manager, TechCrunch+
@yourprotagonist

Interim rate of return: A better approach to valuing early-stage startups

Low valuation caps allow early-stage investors to gain a larger ownership stake and reduce their risk.

However, these caps are increasingly being used as a proxy for the value of the company at the time of the investment, which in turn creates “unnecessary complexity for inexperienced founders and investors,” write attorneys Andrew Ritter, Adam Silverman and Jack Sousa, partners at Wiggin and Dana.

“With the interim rate of return method, you simply negotiate a rate of return (like an interest rate) that applies to the convertible instrument investment solely for purposes of future conversion or the amount payable in a pre-conversion exit.”

3 mistakes to avoid as an emerging manager

Deep tech VC Champ Suthipongchai is a successful fund manager, but he claims to have made plenty of mistakes along the way.

As co-founder and general partner of Creative Ventures, he raised $65 million “with fewer than 25 LPs.” Looking back, he says he initially wasted too much time chasing investors and failed to use FOMO to his advantage.

“While there’s no one right way to go about fundraising, there are a few wrong ways — and failure is a wonderful teacher,” says Suthipongchai.

Lessons for raising $10M without giving up a board seat

Over the last two years, intelligent calendar platform Reclaim.ai raised $10 million “using a more incremental approach,” writes co-founder Henry Shapiro.

“We’ve done all this without giving up a single board seat, and Reclaim employees continue to own over two-thirds of the company’s equity,” rejecting conventional wisdom that founders should “raise as much as you can as fast as you can.”

In a TC+ post, Shapiro reviews the process they used to identify follow-on investors, shares the email template they used to pitch the SAFE and explains why “a larger cap table means more founder control.”

Pitch Deck Teardown: Juro’s $23M Series B deck

Legal tech startup Juro raised a $23 million Series B earlier this year to scale its web-based contract negotiation platform.

Juro’s founders shared their 15-slide pitch deck with TC+ and only “blurred out part of its future road map and the actual numbers for the financials.”

Dear Sophie: Are there any visas or green cards I can get on my own?

Dear Sophie,

I’m so worried and stressed about all the layoffs! I’m safe for now, but it has made me realize I need to take control of my own destiny.

Are there any visas or green cards that I can apply for on my own without relying on my employer?

— Silicon Stressed

4 ways to use e-commerce data to optimize LTV pre- and post-holiday

E-commerce startups make as much as one-fifth of their yearly revenue in the months after Black Friday/Cyber Monday. But how can brands convert shoppers who respond to a holiday promotion into repeat customers who come back all year long?

In a TC+ post, Dan LeBlanc, CEO and co-founder of data and analytics firm Daasity, provides a detailed strategy guide aimed at helping marketers boost ROI and perform cohort analysis to track lifetime value against customer acquisition cost.

“Consumer brands who know how to use their data to maximize LTV will win the holidays and set their brand up for growth well into the new year.”

Top 3 riskiest misconfigurations on the Salesforce platform

No-code technology can be a double-edged sword.

Platforms like Zapier and Salesforce make it easy to automate tasks and workflows, but “configuring a low-code platform is so easy that the low-code administrator often does not understand the impact of checking a box,” writes David Brooks, senior vice president of product at Copado.

In a post for TC+, he breaks down the three riskiest Salesforce misconfigurations:

Modify All Data (MAD) and View All Data (VAD)
Sharing & Sharing Groups
Running Apex code without the “runAs” method

Startup founders need to be data-informed, not just ‘data-driven’

According to Ann Lai, a general partner at Bullpen Capital, many startups that put core metrics front and center during fundraising are sabotaging themselves.

“Using raw, unfiltered data is common at startups that donʼt know how to properly filter their information, and they often end up offloading data irrelevant to their company and mission,” says Lai.

In a post aimed at both investors and founders, Lai offers three strategies that will help “ensure that you arenʼt just data-driven, but data-informed.”

TechCrunch+ roundup: Fundraising under scrutiny, optimizing LTV, visa bulletin update by Walter Thompson originally published on TechCrunch

Cloud tech companies are facing a significant cloud skills shortage, making it hard to hire people and difficult to make sure their current workforce’s skills are up to date. Australia- and US-based Saasguru wants to narrow the gap with an edtech platform designed for new graduates and tech workers who want to become better at using cloud platforms like Salesforce or AWS. The company announced today it has raised a seed round of $4 million AUD (or about $2.7 million USD) led by Square Peg Capital, along with returning investors Black Nova and Antler.

Saasguru’s last funding was nine months ago, when it raised a pre-seed round of $1.3 million AUD. The company was founded in 2021 by Amit Choudhary, Atif Saad and Prateek Kataria. Choudhary and Saad sold their last startup SaaSfocus, a Salesforce consulting company, to Cognizant in in 2018.

So far, Salesguru has been used by 40,000 students in 20 countries, and has worked with 20 cloud consulting companies that want to train new workers, as well as refresh the skills of their existing teams. Its students range from new graduates who are starting their first jobs in cloud tech to professionals who want to earn more training certificates.

The search for people with cloud computing skills in the Asia Pacific region is urgent, with a report by AWS showing that workers needed will triple by 2025, going from 37 million workers in 2020 to 109 million. Saasguru wants to help its learners become ready for cloud tech jobs, while creating more talent at scale.

Choudhary told TechCrunch that the idea for Saasguru was planted while he and Saad were still working on SaaSfocus and struggled to compete for talent with large cloud consulting companies.

“This forced us to look at organic talent creation by hiring people from diverse non-technology backgrounds and upskilling them through a homegrown program tailor-made for Salesforce job readiness,” he said. “This became a bit of a ‘secret sauce’ for us and it helped us scale the business to over 360 consultants, with over 80% of them being trained through this program.”

SaaSfocus’ training program included hyper-personalized study plans, “TikTok-like” micro-modules of content, mentoring, peer-to-peer learning and hands-on assignments.

After selling SaaSfocus, Choudhary and Saad used this approach in a pro-bono program to help people get new jobs during COVID by teaching them Salesforce skills. Of the 50 people who took part in the program, almost all got placed in Salesforce-related jobs.

“It was the lightbulb moment when we realized this could be scaled with tech into a global business,” Choudhary said. Saasguru was launched in early 2021, combining the components of the pro-bono program with a deep tech platform.

Saasguru’s 15 programs includes ones for learning Salesforce, ServiceNow, AWS, GCP and Azure. It plans to use its funding to add more cloud certifications. Choudhary said Saasguru personalizes courses, which can take from 30 hours for self-paced cloud certification program to 300 hours for a career bootcamp, by using a two-step process. The first step is an initial assessment that analyzes the readiness of a learner and creates a learning pathway for them. Then as they start taking a course, the platform recommends the next best step to take.

Saasguru acquires customers by running free webinars with its teachers, or gurus. They also offer free one-on-one mentoring sessions on careers, interview tips and certifications, and run a Slack community. Saasguru serves both individuals and cloud consulting companies that want to build the skills of new and existing employees.

In a statement about the funding, Square Peg Capital principal Lucy Tan said, “There is a massive cloud skills shortage in the industry that is slowing down digital transformation initiatives undertaken by businesses. Universities are not well equipped to solve this skills shortage as the skills update so quickly. This means post-university upskilling is critical for continued business growth and Saasguru provides a personalised learning pathway for cloud professionals to embark on, helping them get skilled and certified in cloud technologies. This can make a meaningful impact on people’s lives from either landing them in a new career or getting salary increases.”

Edtech Saasguru wants to fix the cloud talent shortage at scale by Catherine Shu originally published on TechCrunch

NASA’s plan to decommission and deorbit the International Space Station (ISS) by 2031 creates another collaborative, multinational opportunity to provide a persistent orbital presence. Yet, the path to achieving that goal remains unclear.

Here’s the rub. No one knows exactly what that presence looks like, how it should be built, who would run it and how it would make money. These complex questions are why we’re thrilled that Blue Origin’s Shahir Gerges, director of business strategy for Orbital Reef, will join us for a fireside chat onstage at TC Sessions: Space on December 6.

These aren’t issues for engineers and astronauts to solve, although they must be involved in the discussion. Instead, it will take far-sighted business leaders who see where the market is headed. They’ll also need to build a compelling case for how a phenomenally expensive expedition, like a long-term space station, can reasonably be expected to pay for itself over a 10-year period.

What, who and how much will it take to build a safe, thriving commercial space economy? What can and should the world expect from privately operated successors to the ISS?

In a session called “Space Station Shake-up,” Gerges, an expert in the realm of commercial space stations, will share his perspective. Be in the room to hear what this space ace has to say about that and more.

Shahir Gerges serves as the director of business strategy for Orbital Reef, within Blue Origin. Orbital Reef is designed to be a mixed-use space station in low Earth orbit for commerce, research and tourism by the end of this decade. Focused on long-term financial sustainability for Orbital Reef, Gerges develops offerings to cultivate growth in new and emerging markets that would benefit from the on-orbit environment, including microgravity.

Before joining Blue Origin, Gerges worked as a strategy consultant at PricewaterhouseCoopers, where he advised industrial companies (including aerospace and defense) on market strategy decisions, internal operations strategy and multiple-deal due diligence. Gerges started his career at United Launch Alliance working in various engineering and strategy roles, as well as supporting government affairs.

Gerges holds a bachelor’s degree in aerospace engineering from Illinois Institute of Technology and an MBA from Georgetown University.

TC Sessions: Space takes place on December 6 in Los Angeles. Buy your pass today, join us to learn about the latest space economy trends, see cutting-edge technology, and network for opportunities to help you build a better, stronger startup.

Is your company interested in sponsoring or exhibiting at TC Sessions: Space? Contact our sponsorship sales team byfilling out this form.

Blue Origin’s Shahir Gerges discusses a post-ISS orbital economy at TC Sessions: Space by Lauren Simonds originally published on TechCrunch

AWS introduced Contact Center, its customer service oriented product some years ago, putting it smack dab in the middle of enterprise applications. It also places the company in the position of competing directly with the likes of Salesforce and other established enterprise SaaS vendors.

When you are competing in that space, you need some powerful features, and today at AWS re:invent in Las Vegas, AWS CEO Adam Selipsky introduced three features to help bring more automation to managing Amazon Contact Centers running on AWS.

For starters, the company is introducing new performance management capabilities under Contact Lens for Amazon Connect designed to help managers identify CSAs who are having issues. The solution uses a combination of performance review forms and machine learning-driven voice analytics to review job performance.

In reality, it’s supposed to help identify agents who might need additional training or coaching. “These reduce the time the contact center managers spend identifying performance issues and helping to coach agents,” Selipsky explained today. Employees could see it differently (the bot says I didn’t answer correctly).

Somewhat along the same lines, AWS is also introducing a new capability to guide agents through customer interaction so they can resolve issues faster and in a more consistent manner. This should help reduce the number of mistakes, and the need for the prior feature (at least in theory).

The company also announced the general availability of Amazon Connect forecasting, which was originally announced in March this year. It’s designed to help contact center managers optimize agent schedules and ensure that they have the right people available.

“Connect is a great example of how the cloud is removing constraints to reimagine business challenges like delivering better customer service,” Selipsky said, something that SaaS companies have known all along, but for AWS, which tends to concentrate on infrastructure and platform pieces, it is a different approach.

AWS adds automated agent monitoring to Amazon Contact Center by Ron Miller originally published on TechCrunch

Crypto lending platform BlockFi filed for Chapter 11 bankruptcy on Monday, just a few weeks after once-major crypto exchange FTX did the same. While BlockFi has been struggling to stay afloat for months now (and was even potentially going to be acquired by FTX), this latest filing signals that the bankruptcy contagion may run deeper than what the crypto industry sees at the surface.

“It is another example that the crypto winter is not over, and with the FTX debacle, it’s going to persist longer than previously expected,” Ric Edelman, founder of Digital Assets Council of Financial Professionals (DACFP) and author of “The Truth About Crypto,” said to TechCrunch.

With BlockFi now in the midst of bankruptcy proceedings alongside FTX, Celsius, Three Arrows Capital and Voyager, others in the crypto space are wondering if and when the next crypto firm will find itself on the chopping block.

“The level of intercompany investment within the blockchain and crypto community is unusually high, so the level of contagion is likely to be much higher than what we saw within the traditional financial community in 2008-2009,” Sam Dibble, partner at Baker Botts, said to TechCrunch.

“Unfortunately, the complexity of these intercompany investments is not fully known at this point and so the companies themselves probably don’t know how badly they will be impacted,” Dibble added. “My guess is that the entire industry will be impacted somewhat by the FTX and BlockFi bankruptcies, with a significant number of severe casualties.”

As BlockFi files for bankruptcy, how contagious will FTX’s downfall become? by Jacquelyn Melinek originally published on TechCrunch

Snapchat is adding a new privacy setting that enables users based in California to better protect their sensitive personal information. The company confirmed it’s rolling out a feature designed to comply with the California Privacy Rights Act (CPRA), which takes effect on Jan. 1, 2023, and applies to personal data collected on or after Jan. 1, 2022.

In November 2020, California residents voted to pass the CPRA, also known as Proposition 24, which builds on an earlier consumer privacy law, the California Consumer Privacy Act (CCPA) of 2018.

While the CCPA gave residents the right to access and delete personal information held by businesses and opt out of the sale of that data, the new law puts into place further requirements for businesses around their data collection practices and data retention. It additionally introduces new notification requirements and clarifies that users have the right to opt out of both the sharing and the sale of their personal information, while also adding a new category of “sensitive data.”

The law created the California Privacy Protection Agency to enforce the state’s privacy laws, as well as investigate violations, and assess penalties if warranted.

Consumers based in California, meanwhile, are to gain the right to not only know who’s collecting their information, but also be able to access it, correct it, delete it and transfer it, and to stop its sale and sharing, without being penalized as a result. As part of this, they’re also to gain the ability to access their options through “easily accessible” self-serve tools.

Snapchat’s implementation would seemingly address the latter as it presents a simple toggle switch under its Privacy Controls section in the app’s Settings screen. Here, users will be presented with a new option at the bottom of the list that reads “California Privacy Choices.”

A tap into this screen, (as spotted by competitive intelligence provider Watchful — see below image) reveals a new option to “Limit the Use of Sensitive Personal Information.” This page explains that enabling the setting would require Snapchat to limit the use of users’ personal info, including things like precise geolocation.

The setting, however, is appearing in the Snapchat app for all U.S. users — even those who don’t live in the state.

Snapchat confirmed the new privacy feature is rolling out in compliance with the CPRA, but notes it only functions for those users in California.

The addition is interesting as it demonstrates how a popular mobile app has chosen to comply with the new legislation. And unlike on Facebook — where settings are buried, difficult to find, and constantly being relocated — Snapchat’s new privacy feature is relatively easy to find. All of the app’s settings are available from one main screen, organized into sections. So the new CPRA-compliant setting isn’t something users have to dig around to find.

Snapchat complies with the California Privacy Rights Act with a new toggle switch for users by Sarah Perez originally published on TechCrunch

At its re:Invent conference, AWS today announced Amazon DataZone, a new data management service that can help enterprises catalog, discover, share and — most importantly — govern their data. The nifty part here is that AWS is using machine learning to help businesses build these data catalogs and generate the metadata to make it searchable.

“To unlock the full power the full value of data, we need to make it easy for the right people and applications to find, access and share the right data when they need it — and to keep data safe and secure.” AWS CEO Adam Selipsky said in today’s keynote.

The tool will provide users with fine-graned controls to manage and govern this data. That’s long been a major problem for enterprises, but it has only gotten harder as the amount of data has increased, ensuring that the right users have access to the right data, without compromising personally identifiable information, for example.

“Data zone enables you to set data free throughout the organization safely by making it easy for admins and data stewards to manage and govern access to data,” Selipsky explained. “And it makes it easy for data engineers, data scientists, product managers, analysts and other business users to discover, use and collaborate around that data to drive insights for your businesses.”

DataZone users will get access to a portal where they can set up their data catalog and define the taxonomy. Once DataZone is connected to a data source, it’ll use machine learning to populate its catalog with its metadata and users can add additional labels and descriptions as needed.

In a somewhat related announcement, AWS also yesterday launched its Digital Sovereignty Pledge, promising to give users the tools they need to control where their data is stored and accessed. DataZone does offer some of those controls, though its focus is obviously not on digital sovereignty.

AWS launches DataZone, a new ML-based data management service by Frederic Lardinois originally published on TechCrunch

Lordstown Motors has starting shipping its all-electric Endurance pickup trucks manufactured by Foxconn, a milestone that seemed impossible earlier this year.

Lordstown Motors, which has experienced investigations, executive upheaval and a shortage of capital, said Tuesday that its full-sized EV truck received full homologation with certification from both the EPA and CARB that clears the way for the company to start customer sales.

The first batch of 500 EV pickups, made at an Ohio factory now owned by Taiwanese hardware manufacturing company Foxconn, are on their way to fleet customers, according to the company. The announcement sent shares of Lordstown Motors stock up 3.79%.

Lordstown Motors has moved from one dramatic event to another in its short life as a company. Workhorse Group founder and former CEO Steve Burns started Lordstown Motors in 2018. The startup immediately gained attention for its deal with GM to buy the legacy automaker’s soon-to-be-shuttered factory in Lordstown, Ohio. GM invested $75 million into Lordstown.

The company became a political tool at times; the unveiling of its Endurance pickup truck geared towards contractors was largely used as a campaign stop for then-Vice President Mike Pence. Lordstown also sustained a series of setbacks, which seemed to ramp up after it became a publicly traded company through a merger with a special purpose acquisition company.

In March 2021, Hindenburg Research, the short-seller firm whose report on Nikola Motor led to an SEC investigation and the resignation (and ultimately the indictment) of its founder, issued a report disputing Lordstown’s claims it had booked 100,000 pre-orders for its truck. The firm also alleged that Burns paid consultants for every truck pre-order as early as 2016 while he was leading Workhorse.

By summer 2021, Burns and CFO Julio Rodriguez had resigned, the company warned it was running low on capital and investors learned it was being investigated by the Department of Justice.

Foxconn, best known as the maker of Apple’s iPhone, came to the rescue in September 2021 and bought the Lordstown factory for $230 million. Foxconn also agreed to manufacture the pickup truck for Lordstown under a joint venture contract. Foxconn increased its investment in Lordstown Motors in November 2022 by buying $170 million in common stock and newly created preferred shares.

Foxconn holds all of Lordstown’s outstanding preferred stock and 18.3% of its common stock on a pro forma basis.

Lordstown Motors begins shipping its Foxconn-made EV pickup trucks by Kirsten Korosec originally published on TechCrunch

Amazon announced today the launch of “Create with Alexa,” a new AI tool for kids that generates animated stories. The company first revealed the feature in September. “Create with Alexa” launched in the U.S. today, November 29, across supported Echo Show devices and is available in English.

To craft a story, a child says, “Alexa, make a story,” and then answers prompts, the company explains in its blog post. The child selects from three themes: “space exploration,” “underwater” or “enchanted forest,” and then chooses the story’s hero, a color scheme and adjectives like “silly,” “happy” or “mysterious.”

The AI then generates a five-to-ten-line story based on the answers. The story comes with a background image, animations, sound effects, music and more. Amazon claims that the story will be different every time, even if the same child chooses the same exact prompts.

Users can also save the story in their personal media gallery and replay it whenever they want. Soon, they’ll be able to share the story with their family and friends, Amazon added.

Nico Bishop, a UX designer, said that Alexa’s AI-assisted story creation would evolve over time. “Kids will want to be able to say, ‘I want to see a wall of chocolate,’ and soon we expect we’ll be able to do that,” Bishop said in a statement. She also noted that the company is working to improve graphic latency, sound effects and music capabilities.

“We envision a world where anyone can bring their ideas to life in the form of digital creations just by using a few spoken words,” said Eshan Bhatnagar, head of product for Alexa AI. “And Alexa will be right with them, assisting them as their co-creator.”

Some AI art systems like Stable Diffusion lack safeguards, making it easy for users to generate inappropriate artwork. However, Amazon’s “Create with Alexa” is targeted toward young kids, so there are content filters in place, Bhatnagar explained. And since it’s a child-directed Alexa feature, a parent needs to enable it on the device before their child can use it.

Amazon’s new Alexa feature uses AI to create animated kids’ stories on Echo Show by Lauren Forristal originally published on TechCrunch