Persona’s founders are certain the world can use another humanoid robot


Jerry Pratt and Figure quietly parted ways last month. The MIT research scientist spent just under two years with the Bay Area-based robotics firm. In 2022, he left Boardwalk Robotics, a humanoid startup he founded and led, and joined Figure’s well-funded ranks as CTO months before it exited stealth.

It was only last week, however, that Pratt made his exit public. The news arrived via LinkedIn, as he announced the founding of another entry in the increasingly crowded world of humanoids. Persona AI is presently as early stage as early stage gets, having been officially founded only last month.

The startup is the brainchild of Pratt and longtime associate Nic Radford, an industry vet with his own impressive resume including seven years as part of NASA’s robotics before founding Nauticus Robotics and Jacobi Motors.

“We wanted to get some early indications from both people who wanted to work with us and investors, that if we did something like this on LinkedIn, it wouldn’t fall flat on its face,” Radford told TechCrunch.

The news was as much a hiring announcement as brand unveiling. “Hey LinkedIn!” Pratt enthusiastically noted on the business site. “Ever dreamt of creating your own Iron Man suit but without the billionaire playboy part?”

Radford and Pratt say they want to bring on additional 10 to 20 “founders” (their quotes) to help shape the company. “Jerry and I are obviously a pivotal part of this,” Radford said, “but so will the next 18 people in. We really want to illustrate to them the esprit de corps of the company.”

At this early stage, Persona’s pitch doesn’t stray far from the various humanoid firms with which it’s set to compete. The introductory text on its website is largely a celebration of those technological breakthroughs that form the foundation of this unique moment in robotics.

The founders write,

Now is a good time for the commercialization of humanoids. Computer vision and perception algorithms can now detect motion, identify and segment objects, and estimate poses at frame rate; electronics and computation have shrunk in size and increased in performance, such that they can be fully onboard a robot and not hog the energy budget; mobility and manipulation algorithms are now competent enough to maneuver around rooms and do commercially useful work; machine learning is increasing robot capabilities while reducing programming burden; investors are starting to believe in the potential of humanoids; and commercial entities are requesting humanoid robots in various applications where they can add real value.

That’s about as deep as the pitch currently goes outside of investor decks and employee interviews. Whatever advantage Persona believes it will ultimately have over Agility, Boston Dynamics, Figure and the rest isn’t clear at this very early stage.

“In some ways, it’ll be very similar, in other ways it’s different,” Radford answered cryptically. “It’s like the way GM feels themselves against Ford or Toyota or any car company. Every company feels, deep down, they’ve got certain competitive advantages. And then, deep down, every company’s commoditized and distilled down into the same things. They all provide transportation. Do we have our version of the Dodge Hemi? We’d like to think so.”

Pratt, for one, felt confident enough in Persona’s vision to leave a top spot at one of humanoid robotics’ most prominant and best-funded companies, Figure. Pratt says the split was amicable, and when I spoke with Figure founder and CEO Brett Adcock last week about his new project, Cover, he spoke highly of his former CTO. Pratt says the decision was, in part, geographical.

“I was going between Pensacola [Florida] and California every two weeks,” Pratt said. “At first when I joined Figure, I thought [Pratt and his wife] could move to California at about the two-year mark. I had planned to do it, but it just really wouldn’t work out. It was a fairly mutual parting of the ways.”

Rather than setting up shop in a traditional robotics hotbed like Boston or Pittsburgh, Persona will split its operations between Pratt’s home of Pensacola, along with Houston. The latter will serve as the company’s primary headquarters, eventually accommodating around two-thirds of Persona’s staff.

Anterior grabs $20M from NEA to expedite health insurance approvals with AI


Anterior, a company that uses AI to expedite health insurance approval for medical procedures, has raised a $20 million Series A round at a $95 million post-money valuation led by NEA, according to two people familiar with the deal. Existing investors Sequoia, which led Anterior’s $3.2 million seed round last September, and Neo, an accelerator that helped the company launch in the summer of 2022, also participated in the Series A financing.

The round also included a host of angel investors, including Mustafa Suleyman, a DeepMind and Inflection AI co-founder who was hired by Microsoft in March to lead the tech giant’s consumer AI division.

NEA and Anterior didn’t immediately respond to a request for comment.

Anterior, formerly known as Co:helm, was co-founded by Abdel Mahmoud, a former doctor who left medicine to pursue a master’s degree in computer science and a career in tech after he grew frustrated with the amount of time he spent on administrative functions rather than with patience.   

The company has built an LLM-powered co-pilot that helps nurses and doctors save hours on gathering medical documentation required by insurance. Anterior’s solution aims to reduce denial rates and accelerate patient access to care.

While Anterior’s initial offering is in prior authorization automation, the company eventually plans to expand into other medical administrative functions.

Mohamad Makhzoumi, managing general partner on NEA’s healthcare team and co-CEO of the firm, joined Anterior’s board. Makhzoumi’s investments include Tempus, a genomic testing and data analysis company founded by Groupon founder Eric Lefkofsky, which is planning to IPO next week at a valuation of up to $6.1 billion. Makhzoimi also backed Xaira, an AI drug discovery startup that launched this year with $1 billion in funding.

Anterior competes with Cohere Health, another provider of prior authorization automation, which raised a $50 million round in February led by Deerfield Management, with participation from Define Ventures, Flare Capital Partners, Longitude Capital and Polaris Partners, bringing the five-year-old company’s total funding to $106 million.  

GrubMarket buys Butter to give its food distribution tech an AI boost


Much of how people buy food has moved online — restaurants often replace menus with QR codes that let you order with your smartphones, and grocery shopping has been revolutionized with delivery services like Instacart. But until recently, the other side of the food supply chain — how small restaurants and neighborhood groceries procured food — depended largely on physical media, pen and paper.

Now, GrubMarket, which provides software and services that help link up and manage relationships between food suppliers and their customers, is hoping to make the distribution process more digital and efficient via a new acquisition.

California-based GrubMarket recently acquired Butter, a SaaS platform that aims to digitize the traditionally manual food distribution process with AI, the companies exclusively told TechCrunch. Founded in 2020, Butter’s eight-person team will join GrubMarket, and its software suite will be integrated with GrubMarket’s own slate of offerings.

Mike Xu, founder and CEO of GrubMarket, declined to disclose the price of the deal, but Winston Chi, Butter’s co-founder, told TechCrunch that “most parties, including our investors and us, are making money” from the exit.

Butter’s post-money valuation was $39 million when it raised a $9 million Series A in November 2022, per PitchBook (the company confirmed with TechCrunch the reported valuation is roughly correct). Backed by investors including Google’s AI-focused Gradient Ventures, Uncommon Capital, Notation Capital, Collide Capital, and angel investor Jack Altman, the startup has raised $12.3 million in total.

GrubMarket has been on a buying spree over the past few years and has acquired over 100 companies to date. Most of these deals focus on supply chain consolidation, as the company operates a B2B e-commerce business. On one hand, GrubMarket directly sources produce and ingredients from growers and supplies to buyers like supermarkets. On the other, it sells distributors the software needed to run their businesses. It’s not unlike Amazon’s positioning as both a marketplace and SaaS provider.

Butter, alongside Farmigo and IOT Pay, remains one of the few venture-backed startups in GrubMarket’s portfolio that are aimed at bolstering its tech stack.

It’s unclear whether GrubMarket used capital from its balance sheet for the acquisition. Given its profitability and funding history, it wouldn’t be surprising if the money came out of its pocket — Xu told TechCrunch the company has been profitable on an EBITDA-basis for three consecutive years, and its annual revenue run-rate is on track to surpass $2 billion in 2024.

Xu declined to comment on GrubMarket’s fundraising plans, only saying that it has raised “hundreds of millions of dollars” to date. GrubMarket’s last publicly announced investment happened in 2022, a $120 million round that valued it at more than $2 billion. In late 2021, Bloomberg reported that the company was “interviewing banks” for a potential IPO in 2022.

Scooping up Butter

GrubMarket is effectively buying out a smaller competitor. At the height of the coronavirus pandemic in 2020, Chi and his co-founder, Shangyan Li, launched Butter as an end-to-end vertical SaaS solution to help small and medium-sized food wholesalers manage everything from inventory and customer relationships to ordering.

These aren’t necessarily unique features — GrubMarket itself provides many of them — but like many SaaS startups, Butter quickly jumped on the generative AI bandwagon, developing tools to improve its users’ workflow.

Butter’s voice-to-text feature automatically turns customer voicemails into orders. Image Credits: Butter

The ordering process in the wholesale food industry was particularly ripe for a change. Food suppliers would often scribble orders down as they listened to voicemails from their customers — like a chef calling from a restaurant at the end of the day after counting inventory — or scroll through text messages of orders. This haphazard process often led to wrong orders or missing items. Analyzing sales and performance remained a dream.

Using AI, Butter built features to help distributors turn that type of unstructured data into information that can be viewed, tracked and analyzed easily. It uses a mix of third-party AI models and its proprietary AI to convert voice notes into lists of items that restaurants and supermarkets order. Before the AI-generated information goes into Butter’s system, users get a chance to review it for accuracy. And because the information is now digital, distributors can analyze sales and optimize their inventory and pricing.

“Every sales rep on the distributor side literally spends five hours a day transcribing text messages and voicemail orders, so it’s a huge amount of productivity boost and manual process cut-down,” Li said.

More importantly, Butter doesn’t ask its customers to learn a completely new workflow. “Neither distributors nor restaurants want to change how they communicate. We aren’t changing their workflow, but we are helping them centralize sales knowledge,” said Chi.

“Every single step [of food distribution] can be boosted by AI. Even if we aren’t replacing humans, AI can easily help 10x sales. We start with ordering because this is clearly the biggest pain point,” added Chi.

As it turned out, Butter’s AI capability was the impetus GrubMarket needed to buy and merge with its young rival.

Fast dealmaking is the order of the day

Four years into building Butter, Chi and Li had a sticky product, but they found themselves struggling to scale their customer base without a strong distribution channel.

Looking across the industry, they realized their most formidable competitor, GrubMarket, had the customer reach they needed. They also recognized that Butter could play a complementary role to GrubMarket. Chi and Li decided to propose a merger to Xu.

Butter’s AI assistant helps generate new orders based on text messages. Image: Butter

“The moat is not the tech but the data, and we thought, ‘Wow, GrubMarket has all the data,’” Chi reflected on his decision to sell the company.

Xu had already heard of Butter at the time because the startup had won over a customer from GrubMarket. “[Butter] works harder with the customer […] they even had a team sleeping in the customer’s warehouse to get the job done,” said Xu. “But we all know building an ERP system needs a lot of investment. Winston’s team only raised about $12 million, so it was hard to continue to build a sophisticated ERP system.”

GrubMarket had plans to automate order management, but its development resources were “fully loaded” and focused on other features, like using AI to derive customer intelligence from raw data, according to Xu. So when Butter proposed the deal, the technological synergies were immediately obvious. Furthermore, the startup had a stronghold in a segment that GrubMarket had coveted — seafood distributors. Butter reached out in March, and by the end of April, GrubMarket had already completed the deal to acquire it.

Once the companies have been integrated, GrubMarket will leverage Butter’s products, which include AI-augmented chat commerce, to strengthen GrubAssist, its enterprise AI assistant. GrubMarket is also slated to add an AI-enabled prospecting and digital ordering module to its ERP system, which will let food wholesalers automatically generate digital sales orders regardless of the original medium the orders were taken on — be it text, paper, voicemails, or emails.

“Our style is very direct and fast-moving,” said Xu, commenting on the speed of the dealmaking. “It’s great that [Butter] joins us so we don’t need to build it from scratch, and that’s a great addition to our software product family.”

Scammers found planting online betting ads on Indian government websites


Some Indian government websites have allowed scammers to plant advertisements capable of redirecting visitors to online betting platforms.

TechCrunch discovered around four dozen “gov.in” website links associated with Indian states, including Bihar, Goa, Karnataka, Kerala, Mizoram and Telangana that were redirecting to online betting platforms. Some of those websites belong to state police and property tax departments in the respective states. The scammy links were indexed by search engines, including Google, making the ads easy to find online.

The redirecting websites, touted as “Asia’s most popular” online betting platform and “the number one online cricket betting app in India,” claim to allow betting on games, including cricket tournaments such as the Indian Premier League.

It’s not clear how the scammers planted the ads on Indian government pages or for how long the links were redirecting to the online betting platforms.

Image credits: Google / TechCrunch

After spotting the issue earlier this week, TechCrunch alerted India’s Computer Emergency Response Team, known as CERT-In, to the lapse and provided a few affected state government website links for reference.

Shortly after, the Indian cyber agency acknowledged the receipt of the email, and on Thursday CERT-In confirmed it escalated the matter.

“We have taken up with the concerned authority for appropriate action,” the agency said in an email response. It is not clear if the flaw allowing the backdoor access to state government websites has been fixed.

Last June, TechCrunch reported that scammers had published ads for hacking services on U.S. government websites by way of a security flaw in the government’s web content management system software. Some of those ads appeared to be available online for years.

Tesla layoffs hit high performers, some departments slashed, sources say


Tesla management told employees Monday that the recent layoffs — which gutted some departments by 20% and even hit high performers — were largely due to poor financial performance, a source familiar with the matter told TechCrunch.

The layoffs were announced to staff just a week before Tesla is scheduled to report its first-quarter earnings. The move comes as Tesla has seen its profit margin narrow over the past several quarters, the result of an EV price war that has persisted for at least a year. The company delivered a record 1.81 million vehicles in 2023. Its margins, however, took a hit after Tesla repeatedly slashed prices in a bid to drum up sales and undercut the competition.

Tesla informed employees that more than 10%, or about 14,000 workers, will be laid off across the global organization that has operations in the United States, Europe and China. In a regulatory filing, Tesla referred to the l layoffs as a “company-wide restructuring.” The layoffs, which affected employees across all departments and seniority levels, were made to reduce costs and increase productivity to prepare for its “next phase of growth,” according to an internal email from CEO Elon Musk that TechCrunch has viewed.

High performers also cut

Many of the laid-off employees were high performers, according to two sources who spoke to TechCrunch on condition of anonymity. One source expressed shock at the number of talented employees cut and noted that many of those affected were working on projects that have fallen lower on Tesla’s priority list. The source declined to specify which projects.

Some departments saw layoffs beyond the 10% outlined in the companywide email, according to sources. One manager told TechCrunch that 20% of their employees were cut.

“I lost 20% of my team, some really good players too,” they said.

The shakeup also comes as Musk continues to bend the company’s trajectory toward building fully self-driving cars. Tesla recently dropped plans to build a lower-cost EV that would retail starting at around $25,000, opting instead to use the underlying platform being developed to power an alleged robotaxi that Musk said will debut August 8.

Musk previously tried to prioritize the dedicated robotaxi vehicle project, according to his biographer, Walter Isaacson. In 2022, he told employees that he wanted a “clean robotaxi” with no steering wheel or pedals. Tesla lead designer Franz von Holzhausen and engineering VP Lars Moravy kept running the low-cost EV project in secret and eventually convinced him to make both — that is, until last week when it was reported that Musk changed his mind.

Top execs leave

Two high-profile executives — Drew Baglino, Tesla’s SVP of Powertrain and Energy, and Rohan Patel, VP of Public Policy and Business Development — also left the company.

Patel told TechCrunch he decided Sunday evening to leave Tesla because of “[b]ig overall changes” at the company. Patel, who had been engaging regularly with Tesla customers and fans on X in recent months, declined to be specific. He noted in a message that it would be “Better for me not to speculate.” “Tesla is going to be stronger than ever, and change is good,” he added.

Baglino told TechCrunch that after 18 years it was time to leave Tesla. “I feel good about the impact I’ve been able to achieve, my leadership team is strong, the energy businesses I’m responsible for are doing well, etc.,” he wrote in a message to TechCrunch.

“Baglino was in charge of powerdrives and new battery projects, and there’s a sense that there isn’t a whole lot of innovation that’s sustainable at this point, which is probably why Baglino is leaving,” Sandeep Rao, head of research at London-based financial services company Leverage Shares, theorized in an interview with TechCrunch.

Baglino’s departure comes just a few months after Tesla’s previous CFO, Zachary Kirkhorn, stepped down. In January, Musk posted on X, formerly Twitter, that he would want to have around 25% voting control of Tesla in order to focus more fully on the company, rather than on his other companies, and help the EV-maker become a leader in AI and robotics.

This article was updated to include information from a regulatory filing that refers to the layoffs as a “restructuring.”