Trade My Spin is building a business around used Peloton equipment

Peloton Bike Lifestyle 04

Image Credits: Peloton

Peloton had one of the most turbulent half decades in tech. The home fitness firm experienced some of the industry’s highest highs and lowest lows in dizzying succession. It’s the story of a buzzy startup that garnered a cult-like following among influencers and fitness fanatics. A global pandemic shot the brand to unknown heights, before over-investment, recalls, mass layoffs and executive departures brought the brand crashing back down to earth.

As of mid-2024, Peloton is down but not out. The company avoided a major liquidity crunch with a massive debt refinance at the end of May. That marked the end of a month that also saw a 15% staff reduction and the exit of CEO Barry McCarthy a little over two years after he took over for founder John Foley.

Peloton’s high-profile roller-coaster ride has had wide-ranging knock-on effects. Excitement peaked at the height of the pandemic, but once the world began to reopen, sales cratered. Some who were hooked at the height of social distancing have remained loyal to the brand. Plenty of others, however, lost that connection. A degree of attrition is inevitable with any fitness offering, but those figures were unquestionably exacerbated by the reopening of gyms and other exercise alternatives.

The result is that a lot of unused pieces of pricey fitness equipment are occupying space in homes across America; they’re now “clothing racks,” as a colleague recently referred to her Peloton bike. A quick search on Facebook Marketplace reveals row after row of the stationary bike, routinely listed around $300 to $500 — a fraction of the cost of a new model (around $1,500). For many once enthusiastic owners, the hardware has become a nuisance.

But for a pair of East Coast entrepreneurs, it’s an opportunity.

The Trade My Spin origin story starts modestly, when now CEO Ari Kimmelfeld began looking for a good deal on a used Peloton bike. As good as the Facebook and Craigslist prices were, relative to buying new from the manufacturer, the experience had its own issues.

“There was a massive inconvenience, buying something that bulky,” Kimmelfeld, who was then working at Ernst & Young’s strategy consulting arm, EY-Parthenon, tells TechCrunch. “Five hundred dollars was a lot of money, and meeting up with a stranger and giving them money for a piece of equipment that you can’t really test out. Also, I live in New York City. Getting something like that from an apartment in Brooklyn to Manhattan is difficult. Also, there’s no warranty.”

Local logistics

Image Credits: Trade My Spin

Kimmelfeld began a pilot for what would become Trade My Spin last year, picking up and selling used Peloton equipment. At its heart, the offering was a DIY logistics play, removing the friction from buying and selling used exercise equipment. It was a conversation with Joey Benjamini that transitioned the one-man operation into a viable business.

Benjamini built contractor-based logistics network Collectible Classics. His Pennsylvania-based vintage car dealership relies on those contracted drivers to deliver vehicles primarily sold through used car platform, Bring a Trailer.

“Logistics are the most complicated and the most important part of this business — and the biggest barrier to entry,” Benjamini tells TechCrunch. “We have a database of 1099 contractors who do deliveries for us. We’re constantly growing that network of drivers who know our company and our process. Once a driver is trained, we dispatch them to pick up bikes. It’s very simple.”

The new team began work on the Trade My Spin site prior to seeking funding. The page remains simple, even as the inventory has grown to include Peloton’s treadmills, rower and a variety of accessories. A Buy button displays the service’s bustling marketplace, while Sell surfaces a form for the equipment you’re looking to unload. With the site in place, the young company raised a small pre-seed to scale operations.

Talking to Peloton

The startup has also had multiple conversations with Peloton since officially launching in March. Trade My Spin’s primary goal with the calls is convincing that theirs is a symbiotic — rather than parasitic — relationship. At first glance, it’s easy to understand why Peloton might be antagonistic toward the company.

Viewed as a zero-sum game, every used bike sold represents a potential lost sale on a new bike. While it’s true that keeping bikes in circulation is a net positive on the sustainability front, Peloton shareholders are no doubt looking at the sales bottom line in hopes of seeing a turnaround.

The math changes, however, when considering that Peloton’s ultimate goal is being a content company that sells hardware, rather than the other way around. Rather than simply every used bike sale being a missed sale on a new one, Trade My Spin’s pitch is that every bike removed from circulation is one fewer subscription to Peloton’s content platform of classes.

“Every bike we take is from someone who is not using that bike,” says Benjamini. “If someone’s not using the bike, they’re not using the subscription. Peloton is a subscription service. It’s $44 a month. Every time we flip a bike — and we’ve flipped thousands of bikes — they make $500 a year.”

The relationship would no doubt be different had Peloton been more proactive about moving its own used equipment. Ultimately, however, Trade My Spin stepped in to fill that underserved hole in the market.

A new spin

Trade My Spin Founders (L-R): Ari Kimmelfeld and Joey Benjamini

Trade My Spin has pieced together a logistics network capable of offering same or next-day delivery in most major cities in the continental U.S. More-remote locales can take up to five days to complete, which is still faster than the three to five days Peloton takes to process orders.

In the short term, expansion involves adding more fitness equipment to Trade My Spin’s buying and selling options. Longer term, the company is looking to leverage its growing network of contractors to include the buying and selling of all sorts of unwieldy objects. Trade My Spin will likely require an additional funding round to get there.

“We want to transition,” Benjamini says. “We take it from where we’re currently at, and we build it into a large-scale marketplace for bulky items with logistics. That’s the game plan, and no else is going to do that. There’s a barrier to entry and a moat around the business with regards to having the drivers.”

Parker Conrad, CEO at Rippling talks with Mary Ann Azevedo talk about "Going Global" at TechCrunch Disrupt in San Francisco on October 20, 2022. Image Credit: Haje Kamps / TechCrunch

Parker Conrad says founders have been building software wrong for the last 20 years

Parker Conrad, CEO at Rippling talks with Mary Ann Azevedo talk about "Going Global" at TechCrunch Disrupt in San Francisco on October 20, 2022. Image Credit: Haje Kamps / TechCrunch

Image Credits: Haje Kamps / TechCrunch

What is the right way to build a software business? Many startup advisers say that B2B software should solve one pain point, gain customers, then add features as their company grows. Serial founder Parker Conrad, currently the founder and CEO of Rippling, an HR software startup valued at $13.5 billion in April, thinks that’s the wrong way to do it.

Conrad said on a recent episode of TechCrunch’s Found podcast that he thinks the advice given to software founders over the last two decades has been misguided.

“I think the conventional wisdom for how to build business software has been that you should focus very narrowly, and you should build this one very narrow thing and go very deep,” he said. “I think that as a result of that conventional wisdom, we’ve been building business software wrong for the last 20 years. The side effect of building these very narrow applications is that businesses now have to manage 100 different separate pieces of software to run their business, and there’s a lot of inefficiency in that.”

Conrad is naturally talking his book here. Rippling aims to be an all-encompassing solution for everything from payroll to expense management and corporate cards to IT solutions. When Conrad was asked about how he felt about competition, he responded that it depends what a potential customer is looking for. For finance, Conrad said Rippling competes with Brex and Navan while for payroll solutions it might compete with Gusto.

“The secret to building better business software is building a system where you can build multiple parallel business software applications that are all natively built into the same system,” Conrad said.

He said that building software in this way allows companies to take one dataset and build multiple applications on top of it which can result in a standardized user experience. It also gives companies more options to price their offerings, he added.

While there are definitely areas where it is good to have a company focus and go deep, like cybersecurity, Conrad’s notion that platforms perform better than function-specific companies has merit to it — especially in times of economic depression where enterprises don’t have as much money to spend on software.

TechCrunch reported last year that many SaaS startups were likely going to struggle or have to consolidate after the market froth of 2020 and 2021 resulted in many one-feature startups getting launched and funded.

At the time, Loren Straub, a general partner at Bowery Capital, said something similar. Straub told TechCrunch in January 2023 that no one wants to back a startup producing a product that isn’t much more than one feature because such startups have no moat, or element that sets them apart from potential competitors. She said companies don’t like to spend on single-feature tech in general, let alone when budgets have tightened.

VC Mark Goldberg, a partner at Index Ventures at the time, who has since launched his own fund, echoed that sentiment when he spoke to TechCrunch last year. He said large companies are more likely to use less-than-stellar offerings from their existing contracts before they sign a new one with a one-note startup.

“Before Slack sold to Salesforce, one of the scary things was Microsoft launched Teams,” Goldberg said at the time. “We all in Silicon Valley thought it was a fine product and thought it was better than Slack but a lot of people don’t care. Why do we need a separate vendor? That’s an area where best in class may not have helped.”

Conrad hopes the horizontal approach he’s taken with Rippling both attracts customers and keeps them as the market ebbs and flows. It’s worked so far.

“Now every single software vertical has like a dozen competitors, and it’s extremely crowded,” Conrad said, adding that customers prefer an approach where a software company will “build the system that sort of recombines all of this and gives people one system that combines a lot of these capabilities and wins.”

Trace Machina is building a simulation testing platform to update safety-critical applications

Blue car surrounded by blue circle with shafts of light shining on it to symbolize self-driving cars moving down the road.

Image Credits: Jason marz / Getty Images

When a faulty CrowdStrike update brought down airports, 911 call centers and hospitals last month, it showed how a defective update could impact critical infrastructure. Now imagine that this update was for something like an autonomous vehicle or a warehouse robot, and the implications of a bad update could be even more severe.

Trace Machina, an early-stage startup, is trying to prevent such scenarios with advanced simulation software that enables developers to test updates in a more realistic simulated environment. The company emerged from stealth on Thursday, announcing a $4.7 million seed investment and an open source tool called NativeLink.

CEO and co-founder Marcus Eagan says his company is developing a native, Rust-based system to help test and validate software for autonomous systems like self-driving cars and warehouse automation equipment before these systems are deployed in the real world.

“The way we solve that is by providing a native link between developers and their autonomous vision,” Eagan told TechCrunch. That is precisely why the company’s first product is called NativeLink.

“When developers go from working on web apps to working on robots, it becomes obvious that the existing developer toolkit with Docker, Kubernetes, etc. does not suffice. Engineers need to be able to run experiments and tests on the local hardware directly,” he said.

“NativeLink bridges that gap and provides engineers with a staging environment that enables them to run simulations in resource-constrained environments like an embedded Nvidia GPU chip that are difficult to source for robots, self-driving cars and edge devices.”

Eagan says that previously companies had to build these environments themselves and that limited them to well-funded self-driving car companies or hyper scalers like Google. He wanted to build a system that is as close to the hardware as possible, what he calls “being close to the metal,” and make it accessible to any company.

“There’s a lot of people who’ve gone down this path, but none of them can run with direct hardware access. There’s always been this virtualized layer, this abstraction layer, that frankly made it easier for those companies to build those systems and iterate. We just had to pay the tax of being close to the metal,” he said.

Eagan’s background includes stints at MongoDB, where he helped develop Atlas Vector Search, the company’s first AI product. His co-founder, Nathan Bruer, worked at Google X, the company’s experimental moonshot project center, and also helped build autonomous vehicles at Toyota Institute.

Eagan, who is Black, has had to deal with racism in his career, but he remains focused on building his company, regardless. “I have had to deal with racism and I don’t care. I’m so focused on my goal. Nobody can stop me, nobody can dictate how things are going to go. And I’m very grateful for that from that vantage point because a lot of people who look like me don’t have that freedom,” he said.

He has also had to overcome obstacles beyond racism in his life. He was in a severe car accident when he was a teen that left him critically injured, unable to walk or talk. But he was able to recover, go to college, become an engineer and eventually begin building this startup.

The $4.7 million seed was led by Wellington Management with participation from Samsung Next, Sequoia Capital Scout Fund, Green Bay Ventures and Verissimo Ventures, along with several prominent industry angels.

Parker Conrad says founders have been building software wrong for the last 20 years

Parker Conrad, CEO at Rippling talks with Mary Ann Azevedo talk about "Going Global" at TechCrunch Disrupt in San Francisco on October 20, 2022. Image Credit: Haje Kamps / TechCrunch

Image Credits: Haje Kamps / TechCrunch

What is the right way to build a software business? Many startup advisers say that B2B software should solve one pain point, gain customers, then add features as their company grows. Serial founder Parker Conrad, currently the founder and CEO of Rippling, an HR software startup valued at $13.5 billion in April, thinks that’s the wrong way to do it.

Conrad said on a recent episode of TechCrunch’s Found podcast that he thinks the advice given to software founders over the last two decades has been misguided.

“I think the conventional wisdom for how to build business software has been that you should focus very narrowly, and you should build this one very narrow thing and go very deep,” he said. “I think that as a result of that conventional wisdom, we’ve been building business software wrong for the last 20 years. The side effect of building these very narrow applications is that businesses now have to manage 100 different separate pieces of software to run their business, and there’s a lot of inefficiency in that.”

Conrad is naturally talking his book here. Rippling aims to be an all-encompassing solution for everything from payroll to expense management and corporate cards to IT solutions. When Conrad was asked about how he felt about competition, he responded that it depends what a potential customer is looking for. For finance, Conrad said Rippling competes with Brex and Navan while for payroll solutions it might compete with Gusto.

“The secret to building better business software is building a system where you can build multiple parallel business software applications that are all natively built into the same system,” Conrad said.

He said that building software in this way allows companies to take one dataset and build multiple applications on top of it which can result in a standardized user experience. It also gives companies more options to price their offerings, he added.

While there are definitely areas where it is good to have a company focus and go deep, like cybersecurity, Conrad’s notion that platforms perform better than function-specific companies has merit to it — especially in times of economic depression where enterprises don’t have as much money to spend on software.

TechCrunch reported last year that many SaaS startups were likely going to struggle or have to consolidate after the market froth of 2020 and 2021 resulted in many one-feature startups getting launched and funded.

At the time, Loren Straub, a general partner at Bowery Capital, said something similar. Straub told TechCrunch in January 2023 that no one wants to back a startup producing a product that isn’t much more than one feature because such startups have no moat, or element that sets them apart from potential competitors. She said companies don’t like to spend on single-feature tech in general, let alone when budgets have tightened.

VC Mark Goldberg, a partner at Index Ventures at the time, who has since launched his own fund, echoed that sentiment when he spoke to TechCrunch last year. He said large companies are more likely to use less-than-stellar offerings from their existing contracts before they sign a new one with a one-note startup.

“Before Slack sold to Salesforce, one of the scary things was Microsoft launched Teams,” Goldberg said at the time. “We all in Silicon Valley thought it was a fine product and thought it was better than Slack but a lot of people don’t care. Why do we need a separate vendor? That’s an area where best in class may not have helped.”

Conrad hopes the horizontal approach he’s taken with Rippling both attracts customers and keeps them as the market ebbs and flows. It’s worked so far.

“Now every single software vertical has like a dozen competitors, and it’s extremely crowded,” Conrad said, adding that customers prefer an approach where a software company will “build the system that sort of recombines all of this and gives people one system that combines a lot of these capabilities and wins.”

Hyperspace is building custom instances to accelerate database searches

Data moving through a circuit board with CPU in the center.

Image Credits: Ignatiev / Getty Images

The growth in the demand for generative AI apps has led to a need for larger and larger databases to store the associated data (e.g. model training data). These databases tend to be resource-intensive from a hardware perspective and depending on the algorithms used to orchestrate them, they can be high-latency. Often, companies are forced to make trade-offs between database cost, performance and accuracy.

But it doesn’t have to be this way, says Ohad Levi, the CEO and co-founder of Hyperspace. Hyperspace is building custom cloud instances to accelerate two specific database tasks: lexical searches and vector searches. Lexical searches are a type of keyword-based search that look for exact matches in a database, whereas vector searches consider the semantic meaning and context of the search query.

Levi claims that Hyperspace’s instances, which leverage a combination of FPGAs and GPUs, can deliver up to 10 times faster searches than traditional, non-accelerated databases.

“Our product helps companies dealing with large-scale data retrieval, particularly in AI and generative AI applications,” Levi told TechCrunch. “Unstructured data is outpacing traditional search capabilities. Data retrieval solutions must meet lexical and vector search datasets to meet current market demands.”

Prior to launching Hyperspace, Levi was an optimization engineer at Intel and then a product marketing lead at HP. He says he became frustrated with the limitations of legacy search solutions working for Big Tech, which led him to partner with ex-Intel design consultant Max Nigiri to found Hyperspace.

Hyperspace doesn’t sell its instances. Instead, it sells access to managed database software running in those instances (hosted on AWS for now). Hyperspace’s databases can handle various types of structured and unstructured data, including videos, images and text, and are priced according to size and query volume.

“Hyperspace is a cloud-native managed database that works as a software-as-a-service model, priced per usage,” Levi explained. “Our team is able to design customized AI infrastructure solutions to help enterprises solve their search challenges.”

Hyperspace’s performance gains are impressive if true; Levi says that the company’s instances also deliver 5x higher throughput at a 50% lower cost than a typical database. (Those are average results; Levi declined to offer up a direct comparison with a competitor.) But can Hyperspace convince companies to use a newcomer database platform when there are so many incumbents — like Azure, AWS and Google Cloud — to choose from?

Levi says yes, and he claims that Hyperspace is already seeing some early customer traction. The Tel Aviv-based firm has inked deals with enterprises in the fraud prevention and e-commerce spaces, including Forter, Nsure and Renovai, and tripled its annual recurring revenue and total contract volume over the last year.

Hyperspace also recently closed a $9.5 million seed funding round led by Mizmaa with participation from JVP and toDay Ventures. Levi says that the money will be put toward scaling up Hyperspace’s database offering to “thousands” of instances and launching a free, entry-level plan.

“Hyperspace has an entire pipeline of new innovative products that will drive the search market forward and support the needs of our enterprise and small- and medium-sized clients,” Levi said. “We’re not seeing any headwinds. Every generative AI system is a search system, and search is becoming harder than before. The need for better AI infrastructure is growing daily, and with more data, the need for better search applications is becoming more apparent.”

Peloton Bike Lifestyle 04

Trade My Spin is building a business around used Peloton equipment

Peloton Bike Lifestyle 04

Image Credits: Peloton

Peloton had one of the most turbulent half decades in tech. The home fitness firm experienced some of the industry’s highest highs and lowest lows in dizzying succession. It’s the story of buzzy startup that became the target of a cult-like following among influencers and fitness fanatics. A global pandemic shot the brand to unknown heights, before over-investment, recalls, mass layoffs and executive departures brought the brand crashing back down to earth.

As of mid-2024, Peloton is down, but not out. The company avoided a major liquidity crunch with a massive debt refinance at the end of May. That marked the end of a month that also saw a 15% staff reduction and the exit of CEO Barry McCarthy a little over two years after he took over for founder John Foley.

Peloton’s high-profile roller-coaster ride has had wide-ranging knock-on effects. Excitement peaked at the height of the pandemic, but once the world began to reopen, sales cratered. Some who were hooked at the height of social distancing have remained loyal to the brand. Plenty of others, however, lost that connection. A degree of attrition is inevitable with any fitness offering, but those figures were unquestionably exacerbated by the reopening of gyms and other exercise alternatives.

The result is that a lot of unused pieces of pricey fitness equipment are occupying space in homes across America; they’re now “clothing racks,” as a colleague recently referred to her Peloton bike. A quick search on Facebook Marketplace reveals row after row of the stationary bike, routinely listed around $300 to $500 — a fraction of the cost of a new model (around $1,500). For many once enthusiastic owners, the hardware has become a nuisance. But for a pair of East Coast entrepreneurs, it’s an opportunity.

The Trade My Spin origin story starts modestly, when now CEO Ari Kimmelfeld began looking for a good deal on a used Peleton bike. As good as the Facebook and Craigslist prices were, relative to buying new from the manufacture, the experience had its own issues.

“There was a massive inconvenience, buying something that bulky,” Kimmelfeld, who was then working at Ernst & Young’s strategy consulting arm, EY-Parthenon, tells TechCrunch. “Five-hundred dollars was a lot of money, and meeting up with a stranger and giving them money for a piece of equipment that you can’t really test out. Also, I live in New York City. Getting something like that from an apartment in Brooklyn to Manhattan is difficult. Also, there’s no warranty.”

Local logistics

Trade My Spin

Kimmelfeld began a pilot for what would become Trade My Spin last year, picking up and selling used Peloton equipment. At its heart, the offering was a DIY logistics play, removing the friction from buying and selling used exercise equipment. It was a conversation with Joey Benjamini that transitioned the one-man operation into a viable business.

Benjamini built contractor-based logistics network for Collectible Classics. His Pennsylvania-based vintage car dealership relies on those contracted drivers to deliver vehicles primarily sold through used car platform, Bring a Trailer.

“Logistics are the most complicated and the most important part of this business — and the biggest barrier to entry,” Benjamini tells TechCrunch. “We have a database of 1099 contractors who do deliveries for us. We’re constantly growing that network of drivers who know our company and our process. Once a driver is trained, we dispatch them to pick up bikes. It’s very simple.”

The new team began work on the Trade My Spin site prior to seeking funding. The page remains simple, even as the inventory has grown to include Peloton’s treadmills, rower and a variety of accessories. A Buy button displays the service’s bustling marketplace, while Sell surfaces a form for the equipment you’re looking to unload. With the site in place, the young company raised a small pre-seed to scale operations.

Talking to Peloton

The startup has also had multiple conversations with Peloton since officially launching in March. Trade My Spin’s primary goal with the calls is convincing that theirs is a symbiotic – rather than parasitic – relationship. At first glance, it’s easy to understand why Peloton might be antagonistic toward the company.

Viewed as a zero-sum game, every used bike sold represents a potential lost sale on a new bike. While it’s true that keeping bikes in circulation is a net positive on the sustainability front, Peloton shareholders are no doubt looking at the sales bottom line in hopes of seeing a turnaround.

The math changes, however, when considering that Peloton’s ultimate goal is being a content company that sells hardware, rather than the other way around. Rather than simply ever used bike sale as a missed sale on a new one, Trade My Spin’s pitch is that every bike removed from circulation is one fewer subscription to Peloton’s content platform of classes.

“Every bike we take is from someone who is not using that bike,” says Benjamini. “If someone’s not using the bike, they’re not using the subscription. Peloton is a subscription service. It’s $44 a month. Every time we flip a bike – and we’ve flipped thousands of bikes – they make $500 a year.”

The relationship would no doubt be different had Peloton been more proactive about moving its own used equipment. Ultimately, however, Trade My Spin stepped in to fill that underserved hole in the market.

A new spin

Trade My Spin Founders (L-R): Ari Kimmelfeld, Joey Benjamini

Trade My Spin has pieced together a logistics network capable of offering same or next-day delivery in most major cities in the continental U.S. More-remote locales can take up to five days to complete, which is still faster than the three to five days Peloton takes to process orders.

In the short term, expansion involves adding more fitness equipment to Trade My Spin’s buying and selling options. Longer term, the company is looking to leverage its growing network of contractors to include the buying and selling of all sorts of unwieldy objects. Trade My Spin will likely require an additional funding round to get there.

“We want to transition,” Benjamini says. “We take it from where we’re currently at, and we build it into a large-scale marketplace for bulky items with logistics. That’s the game plan, and no else is going to do that. There’s a barrier to entry and a moat around the business with regards to having the drivers.”

Tezi is building an AI agent for hiring managers

HR concept of robot hand selecting a candidate from a bunch of rectangles.

Image Credits: fatido / Getty Images

AI agents are all the rage right now, and Tezi, an early stage startup is working on one to help HR teams find the perfect candidates for a job opening. The startup claims this bot will sift through resumes to find the ones that match the hiring criteria, find time on the recruiter’s calendar to set up an interview and send out the email to the candidate.

Today the company announced a $9 million seed to help fuel its journey to generally available product.

For now, they are rolling out the alpha product with a handful of design customers just this week, but that’s the vision, according to CEO and co-founder Raghavendra Prabhu. He acknowledged that HR has been using automated resume screening for some time, but Tezi saw an opportunity with the new generation of large language models (LLMs ) to build a more sophisticated recruiting tool for HR.

“I think it’s the combination of reasoning and natural language that we felt gave us an option to build something very, very different from what’s historically been done by software in this space,” he said.

His co-founder and COO, Jason James says that existing tools don’t go far enough in his view. “Let’s say you get a thousand applications for a job. AI or ML or algorithms in the past would be good at saying these resumes are very good,” he said. “But a human still needs to send emails and schedule interviews and all of that. And what’s possible now is an end-to-end workflow, not just basic ranking.”

The founders acknowledged that at this stage, humans need to stay engaged in the process and the hope is that it will be fully automated as models improve. What’s more, the pool of candidates that emerge from any job search is going to be dependent on the quality of the prompts and job descriptions.

Tezi interface showing HR bot interacting with a recruiter.

While they understand that automation can lead to bias, they are working on mitigating that to the extent possible. From their perspective, they are taking whatever inputs come from the hiring manager and assessing that against the resumes in an objective manner. They can’t control what the inputs look like, but they say they are trying to minimize bias on their end.

“If you’re assuming that there’s bias coming in from the employer, we at this time, aren’t going to be excellent at preventing that. What we will be doing on our side is protecting against us adding any sort of bias into the mix through algorithmic means,” James said. They are avoiding looking at historic hiring patterns. They want the models to match by skills and other criteria set by the hiring manager.

They have trained their models on 250 million profiles that they have licensed from data providers, and have been working with OpenAI and Anthropic models so far, and tuning them to their hiring requirements. 

The company is just starting. It launched at the beginning of this year. They are beginning work with 15-20 design customers, and the hope is that they will work out all the kinks and get to a wider beta distribution later this year.

The $9 million seed was led by 8VC and Audacious Ventures with participation from Liquid 2, Afore, Prime Set, South Park Commons and industry angels.

TechCrunch Space: Building (and testing) for the future

Image Credits: TechCrunch

Hello and welcome back to TechCrunch Space. It’s becoming a habit to open each TechCrunch Space newsletter with a bit of an update on Boeing’s Starliner mission, so bear with me. Per NASA officials, the spacecraft will now return its two-person crew to Earth no earlier than June 26, instead of the originally planned date of June 14. In other words, they’ll be spending at least 20 days on board the International Space Station instead of just eight. 

Read my story from last week on the causes of the additional delay and what it might mean for Boeing’s Starliner program.

Want to reach out with a tip? Email Aria at [email protected] or send me a message on Signal at 512-937-3988. You also can send a note to the whole TechCrunch crew at [email protected]For more secure communications, click here to contact us, which includes SecureDrop instructions and links to encrypted messaging apps.

Story of the week

The space industry is all abuzz about how SpaceX’s Starship, Blue Origin’s New Glenn and other heavy-lift rockets will change just about everything. One likely consequence is that spacecraft will get bigger — much bigger — as engineers work outside the constraints of low mass requirements. 

There’s one problem: The current testing regime for spacecraft is focused on payloads four meters or less across. Gravitics and NASA are looking to change that, with a new agreement aimed at addressing this dearth of testing and qualification methods for larger spacecraft. 

Image Credits: Gravitics (opens in a new window)

Scoop of the week

Like many highly valued startups, SpaceX sometimes allows its employees to cash out some of their shares by selling to company-authorized outside investors.

TechCrunch has gotten a peek at an internal SpaceX document about such a tender offer from May 2022. Musk posted on X last month that SpaceX holds such sales for employees about every six months.

These documents offer interesting insights into the investors who are authorized to buy these secondary shares, and the good deals they get. Click the link above to take a look.

South African businessman Elon Musk arrives at the Tenth Breakthrough Prize Ceremony at the Academy Museum of Motion Pictures in Los Angeles, California, on April 13, 2024. (Photo by ETIENNE LAURENT/AFP via Getty Images)
South African businessman Elon Musk arrives at the Tenth Breakthrough Prize Ceremony at the Academy Museum of Motion Pictures in Los Angeles, California, on April 13, 2024.
Image Credits: ETIENNE LAURENT/AFP / Getty Images

Launch of the week

Congratulations to Rocket Lab for nailing its 50th Electron rocket launch! To commemorate the massive milestone, the company posted this pretty poignant video on X that sums up just how far Rocket Lab has come.

What we’re reading

Did you know…that we posted the agenda for this year’s space programming at TechCrunch Disrupt? We are incredibly excited by this year’s lineup, which includes some of the top founders and investors operating in the space industry. Plus fireside chats with none other than Rocket Lab’s Peter Beck and Bridgit Mendler of Northwood Space. Click the link above to learn more.

This week in space history

In last week’s ‘This week in space history’ column, we detailed the flight of Sally Ride, the first American woman to go to space. This week we’re commemorating her return. On June 24, 1983, she concluded her historic trip when the Space Shuttle Challenger touched down in California.

Image Credits: NASA

Sift is building a better platform for analyzing hardware telemetry data

Sift founders Austin Spiegel and Karthik Gollapudi

Image Credits: Sift (opens in a new window)

Less than a year after closing its seed round, software-for-hardware startup Sift announced a $17.5 million Series A led by Google’s venture capital arm GV to scale their platform for analyzing real-time data from hardware systems.

The company is developing a platform that provides a single source of truth for telemetry data. Such data is essential for engineers to understand a machine’s performance; even tiny anomalies, if missed, can spiral into catastrophe. One timely example that Sift provides is the uncrewed Starliner test mission in 2019, which experienced a software error that sent the spacecraft into the wrong orbit entirely and led to further delays and mounting expenses in the spacecraft program. 

Such errors could be avoided with a more comprehensive, yet simplified, software stack for telemetry data, Sift suggests. As opposed to the fragmented sensor data that must be managed by entire teams, or else stitched together with ad hoc solutions, the company is offering nothing less than what it calls “a new paradigm”: a single platform that unifies hardware sensor data ingestion, storage, and review. 

Automation is one of Sift’s biggest differentiators. In the past, a customer may have manually run tests and checked dashboards to ensure hardware health, but with Sift, they can encode “rules” into the platform instead. Sift evaluates those rules against simulations, tests, and operations, and only flags an engineer for data review when it discovers an anomaly. 

“Dashboards are fundamentally the wrong solution for in-depth data analysis because there is too much noise for a human to find the signal,” Sift co-founders Austin Spiegel and Karthik Gollapudi explained in an email.

Over the next 12 months, Sift is aiming to boost every part of the software stack with artificial intelligence, from more robust anomaly detection to data review. The company is also looking to further automate parts of the compliance and regulatory review process, as these certification workflows will help engineers communicate their readiness to regulators and cut through red tape faster, Sift says.

The company already has a list of customers, including many well-known space and hardware startups, like K2 Space, Astranis, and True Anomaly. Spiegel and Gollapudi, two ex-SpaceX engineers, said that “a startup’s need for speed, flexibility and competitive advantage drives it to adopt new tech quickly, such as Sift.”  

“Building internal tools requires a dedicated headcount to create, manage, and maintain and takes years,” they added. “Startups are focused on building their business as fast as possible, so building internal tools — something they may be able to do — is not a priority. Engineering hires and priorities are focused on their product. Hiring engineers with domain expertise to build a highly scalable data storage and analytics solution is challenging.” 

Sift currently has 16 full-time employees and expects to more than double that number over the next 12 months. 

Orby is building AI agents for the enterprise

Image Credits: Getty Images

AI “agents” are generative AI models that can perform actions autonomously, like copying info from an email and pasting it into a spreadsheet, and have been hailed as productivity superchargers. That might be a bit premature, given models’ tendency to make mistakes. But at least a few founders (and analysts and investors) seem convinced that agents are the next frontier in generative AI.

Bella Liu and William Lu are two such founders. Their company, Orby AI, is building a generative AI platform that attempts to automate a range of different business workflows, including workflows that involve data entry, documents processing and forms validation.

Lots of startups offer tools to automate repetitive, monotonous back-office business processes (see Parabola, Tines, Sam Altman-backed Induced AI and Tektonic AI, to name a few). Incumbents, too, like Automation Anywhere and UiPath, have moved to embrace AI to try to maintain pace with the generative AI competition.

But Liu and Lu claim that Orby’s tech stands out for its ability to learn and act on workflows in real time and to understand the patterns and relationships within an enterprise’s unstructured data.

“Orby’s platform observes how workers do their work in order to automatically create automations for complex tasks that require some level of reasoning and understanding,” Liu, Orby’s CEO, explained. “An AI agent installed on a worker’s computer effectively watches, learns and generates automations, adapting the model as it learns more.”

With Orby, which launched out of stealth in 2023, Liu and Lu say that they sought to create AI that could understand some of the low-level decisions being made by workers and abstract those decisions away, freeing up workers to focus on headier things.

Liu previously led AI and automation efforts at IBM, including product planning and AI-related mergers and acquisitions, and was UiPath’s director of AI product management. Lu is a former Nvidia systems engineer who joined Google Cloud as an engineering lead, helping to design generative AI document and database extraction tech.

Orby’s purported secret sauce is a cloud-based generative AI model that’s fine-tuned to complete customer tasks, such as validating expense reports. The model relies partly on symbolic AI, a form of AI that leverages rules, such as mathematical theorems, to infer solutions to problems.

Orby
Orby’s generative AI observes tasks performed by people, then learns to automate these tasks.
Image Credits: Orby

Symbolic AI alone can be inflexible and slow, especially when dealing with large and complicated datasets. It needs clearly defined knowledge and context to perform well. But recent research has shown that it can be scalable when paired with traditional AI model architectures.

“For the last two years, we’ve been engineering this AI model, and have performed successful trials,” Liu said. “There are few pure-play generative AI companies attacking the enterprise head-on with something end-to-end. We are one.”

Liu says that Orby’s model can intelligently adapt to changes in workflows, like when an app’s UI gets an update, by analyzing API interactions and a worker’s browser usage. Having software monitor an employee’s every move sound like a privacy disaster waiting to happen. But Liu claims that Orby doesn’t store most customer data; it only uses certain telemetry data to improve its model, encrypting the data both in transit and at rest.

“Humans are kept completely in the feedback loop,” she added.

Orby, which recently raised $30 million in a Series A funding round co-led by New Enterprise Associates, WndrCo and Wing (sources say at a post-money valuation of $120 million), is competing in a challenging sector. Forthcoming agentic AI from generative AI powerhouses such as OpenAI and Anthropic have dampened the prospects of incumbents and smaller players alike.

Adept, a startup building AI agents technology focused on enterprise applications, is reportedly on the cusp of an acqui-hire deal with Microsoft before it manages to ship a single product. Amazon and Google have released AI agent tooling to little fanfare. Elsewhere, UiPath — despite its ramping up of generative AI initiatives in the past year — saw sales plummet in its most recent fiscal quarter.

Liu says that Orby can come out ahead by taking a systematic go-to-market approach. The company is already generating revenue from around a dozen customers, she says, and plans to put its $35 million war chest toward expanding its Mountain View-based, roughly 30-person team.

“The funds are being used to scale our go-to-market, customer support, product and technical orgs,” she said. “The enterprise market has an insatiable appetite for generative AI solutions that demonstrably improve business performance; they are just trying to figure out where to best apply the technology in the near term before they scale it across their business.”