If you haven’t noticed, a growing amount of code that’s being generated today is “AI-assisted.” In fact, Scott Guthrie, Microsoft’s executive vice president of Cloud and AI, estimated back in March that upwards of 40% of the code that developers were uploading to the AI developer tool GitHub Copilot was both “AI-generated and unmodified.”
Now, the trend is giving rise to startups that promise to keep AI-augmented code from mucking up the works — and investors are taking notice.
Earlier this week, an Israel-based startup, Digma, announced $6 million in seed funding for a continuous feedback platform that runs locally on developers’ machines and helps them analyze their code — including generative AI-created code — to identify issues. Yesterday, a San Francisco-based testing platform called Kolena announced its own funding — $15 million — to build tools to test, benchmark and validate the performance of AI models.
Today, a months-old, four-person, Bay Area-startup called Braintrust is taking the wraps off its own fresh funding round of $3 million. According to co-founder and CEO Ankur Goyal, Braintrust is like an “operating system for engineers building AI software,” one that helps them avoid bad results from AI models. Developers building customer support chatbots, for example, might use Braintrust’s tech to ensure that their chatbot answers questions accurately rather than hallucinating false information.
Like many startups promising the ability to build more reliable AI software, Braintrust has savvy backers. Renowned angel investor Elad Gil is among its investors and helped incubate Braintrust’s initial product. (Gil flagged the round for us, calling the six-week-old outfit “a good one.”) Others of its notable investors include Adam D’Angelo of Quora, Clem Delangue of the buzzy AI outfit HuggingFace and OpenAI co-founder Greg Brockman.
Whether an impressive investing syndicate can help push Braintrust to the front of the pack is an open question. In the meantime, ensuring that AI code doesn’t break a company’s workflow is something Goyal says he was practically born to solve.
The child of doctors, Goyal grew up in Pittsburgh and thought he’d become a doctor, too. “Super nerdy” as a teenager, he says a linear algebra class in high school where he learned about Google’s PageRank algorithm would change his life. (“I get goosebumps just talking about it,” he says.) He moved on from biology, studied computer science at Carnegie Mellon University, then “out of extreme boredom” dropped out his junior year to build a relational database system at MemSQL, an early Y Combinator alum. More than five years later, Goyal co-founded his own company, Impira, and when Figma acquired the company late last year, Goyal became the head of its machine learning platform.
It was a good gig. It also gave Goyal even more insight into the growing challenge of building high-quality software products in this new age of AI everything. So late this past summer, he left to start Braintrust.
“I spent quite some time building old-school software,” he says, “and what’s really different about AI is that it’s inherently non deterministic, meaning if you write code, you can’t really guarantee that it’s going to work. You have to test it on real-world examples.” The process is called evaluation or, colloquially, “evals” and not all companies have high-enough quality data to do enough testing. It’s why Braintrust — which has yet to commercialize its product — is working with companies that do, including the workflow automation company Zapier and the spreadsheet tool company Coda, which are currently beta testing what Braintrust has built.
“Their challenge,” says Goyal, “is ‘Okay, we actually have tons of data, and we have lots of users using our product. But it’s really hard for us to boil that down into a representative set of examples that we can use to test our software.’” With BrainTrust, he says, “They can dump as much data as they want into our product, run evaluations against it, and we’ll help them curate ‘golden datasets’ that they can accumulate over time and use as a measure of whether their software is working or not.”
As a bonus, says Goyal, “We actually run inside of their cloud environments,” which enables Braintrust to operate around thorny compliance issues that could otherwise slow down its adoption within enterprises.
It’s early days, of course, and competition will only grow fiercer in the coming months and years. Deepchecks, an Israeli startup whose tagline is “continuous validation for AI,” is yet another outfit that recently raised seed funding.
Still, Goyal describes Braintrust as exactly the product he needed at Figma, and which didn’t live in the world until Braintrust recently created it. “There’s a whole universe around continuous integration that has developed over the past decade. And that’s kind of turned this into a science of shipping software. But in AI land, that methodology and workflow — until our product — just didn’t really exist.”
Pictured above, from left to right: Coleen Baik (founding designer), Ankur Goyal (CEO), Manu Goyal (founding engineer) and David Song (product manager; part of Elad Gil’s team).
We might be watching a historic moon landing today.
Intuitive Machines’ IM-1 spacecraft, the uncrewed Odysseus, could potentially land on the surface of the Moon on Feb. 22 at around 4:24 p.m. ET, after an eight-day journey through space. While we can’t be completely sure that the landing will be successful, on Wednesday Intuitive Machines said Odysseus “continues to be in excellent health in lunar orbit.”
The spacecraft is expected to land near Malapert A crater in the south pole region of the moon.
Odysseus launched on Feb. 15 on a SpaceX Falcon 9 rocket from Launch Complex 39A at NASA’s Kennedy Space Center. Its landing would make the first commercial moon landing and is quite the feat. Landing on the moon is an infamously difficult task for a variety of reasons, including the lack of GPS systems and atmospheric drag. Only five countries, including the former Soviet Union, the U.S., China, India, and Japan, have landed on the moon without a significant wreck.
A small American company’s robotic spacecraft has brought the United States back to the surface of the moon for the first time in more than a half-century.
Intuitive Machines, a Houston-based space company, landed on Thursday, becoming the first commercial company to reach the moon intact. The unprecedented achievement is a win for NASA, which has invested $2.6 billion in contracts with Intuitive Machines and several other vendors to deliver instruments to the moon over the next four years.
It wasn’t all smooth sailing. A few hours before the landing, flight controllers discovered the spacecraft’s laser rangefinders, which help it avoid hazards on the ground, weren’t working. The team decided to take one more lap around the moon, which bought engineers a couple more hours to troubleshoot the problem. During that orbit, they uploaded a software patch to use onboard NASA lasers, which hadn’t previously been tested in space.
Then, there were some communication challenges, but NASA was quick to call the landing a success, even before a photo was beamed back to Earth.
“Today, for the first time in the history of humanity, a commercial company — an American company — launched and led the voyage up there,” said NASA administrator Bill Nelson in a pre-recorded message during the broadcast. “Today is a day that shows the power and promise of NASA’s commercial partnerships.”
The moon lander dubbed Odysseus touched down on Malapert A crater, about 200 miles from the lunar south pole, just before 6:30 p.m. ET. Many nations and private ventures have set their sights on the region because of its ice, thought to be buried in the polar craters. The natural resource is coveted because it could supply drinking water, air, and rocket fuel for future missions, ushering a new era in spaceflight.
Thomas Zurbuchen, NASA’s former head of science, once described each of the first CLPS endeavors as “taking a shot on goal.” The sports analogy means not every attempt will be victorious, but overall the program will give NASA a lot of chances to achieve its moon-to-Mars goals. By outsourcing NASA’s lunar deliveries — rather than fully owning each mission — the agency believes it will save money. The contract with Intuitive Machines for this mission was $118 million.
“We don’t know how many of the early attempts will be successful,” said Joel Kearns, NASA science’s deputy associate administrator for exploration, during a news conference in November. “But I can tell you that these American companies are technically strong and rigorous, savvy, they’re resourceful, and they’re driven to be successful. They want to secure that first mover advantage in generating this new lunar economy.”
But observers have questioned how cost-effective the initiative will truly be, given the riskiness of flying on inexperienced spacelines. In January, Astrobotic Technologies, the first of the CLPS vendors, tried to get to the moon, but never reached lunar orbit due to a detrimental fuel leak. NASA spent $108 million on that mission and lost five payloads in the process.
“If we’re flying missions at one-tenth of the cost of a NASA mission, and we fail two of them, we still get eight missions for that same price,” Kearns said in a pre-recorded statement during the landing broadcast. “Even with one or two or three failures, this is still a very economical proposition.”
The likelihood of success, especially for novice space programs, is still slim. Historically, less than half of all missions to land on the moon have arrived without crashing. The lunar exosphere — an extremely thin atmosphere of gasses barely held by the moon’s gravity — provides virtually no drag to slow a spacecraft down as it approaches the ground. Furthermore, there are no GPS systems on the moon to help guide a craft to its landing spot. Engineers have to compensate for these shortcomings from 239,000 miles away.
Over the past five years, the private sector has tried and failed. An Israeli nonprofit and company collaborated in 2019 on the so-called Beresheet moon mission, which crashed on the lunar surface after an orientation component malfunctioned. Last April, Japanese startup ispace ran out of fuel on its descent and ultimately crashed. Astrobotic’s Peregrine lander never made it that far and ultimately broke apart as it crashed back to Earth.
But Intuitive Machines’ landing could instill confidence in the burgeoning lunar economy.
“I know this was a nail biter, but we are on the surface, and we are transmitting,” said Stephen Altemus, Intuitive Machines’ CEO. “Welcome to the moon.”
The adoption of open banking and instant payments is moving slowly in the United States compared to other markets around the world, for example, Brazil. That said, the new program FedNow went live in July 2023, and data-sharing regulations are forthcoming, so more potential is on the horizon.
Until then, the co-founders of Zūm Rails say the experiences consumers have with payments continues to be fragmented, meaning companies have to create a tech stack to provide a wide range of services to their customers. The Montreal-based company is taking the approach of providing an all-in-one payments gateway that merges open banking with instant payments.
Marc Milewski and Miles Schwartz started the company in 2019. Milewski’s background is in treasury payments and he was an early employee at accounts receivable automation software company Versapay. While there, he worked on what ultimately became Canada’s first webhook-enabled EFT gateway.
“You learn about all the problems everyone has moving money,” Milewski told TechCrunch. “Open banking was discussed, but I thought it was more about payments. Miles and I talked about building a whole new gateway that unified these experiences. Companies don’t want to be payment experts — that’s our job.”
They started building software to simplify the complexity of moving money via different payment rails so companies can use whichever approach makes sense for their business. Their technology leverages “omni rails” for payments, whether it is traditional credit, debit or electronic funds transfer options. It also provides for real-time options through partners, including Visa Direct, Mastercard, MX and Canada’s Interac network.
Zūm Rails manages the flow of money, including the reduction of fraud and failed transactions, by verifying a customer’s identity, linking directly with bank accounts and facilitating payments via the method of the customer’s choosing.
The company now processes more than $1 billion in payments through its platform each month for over 500 companies, including Questrade, Coinsquare and Desjardins, which is a large federation of credit unions in North America. In the past year, the company grew over 200% and launched in the U.S. at the end of 2023.
Milewski and Schwartz bootstrapped Zūm Rails, building it up to a team of 30 people. Last year, the pair decided to raise venture capital.
“We reached the point where we realize that bootstrapping is no longer healthy for our business,” Schwartz told TechCrunch. “We have some big initiatives we want to work on and grow on. Now it makes sense to do it all at once, and it’s healthy for the business to now go all-in and use the fuel.”
Zūm Rails’ technology leverages “omni rails” for payments, whether it is traditional credit, debit or electronic funds transfer options. Image Credits: Zūm Rails
They closed on a $10.5 million Series A funding round, led by Arthur Ventures, and intend to invest in growing in the U.S. and expanding its payments offerings that will include the introduction of new banking-as-a-service features for merchants. In addition, Zūm Rails is working on a FedNow offering in the U.S. that will enable businesses to send and receive FDIC-insured payments within seconds.
Zūm Rails’ performance to date “is really impressive,” Jake Olson, vice president at Arthur Ventures, told TechCrunch. He called the company “a great fit” for its investment thesis, which is high-growth and capital-efficient B2B software companies.
“Achieving profitability without any outside capital is impressive,” Olson said. “Their product positioning is also really compelling. Rather than weaving together different systems, Zūm Rails can provide organizations with a comprehensive solution that powers the entire transaction journey and enables them to have a seamless experience for their end users. Any organization that views the streamline digital financial interaction coupled with the instant payments capability as a competitive advantage will be a great fit for Zūm Rails.”