|
|
|
|
|
|
|
|
|
|
Why VCs Are Scrutinizing the Revenue at AI Startups
|
|
By Marc Vartabedian, WSJ Pro
|
|
|
|
|
|
|
Good day. In the dizzying world of artificial intelligence startups, figuring out even the most basic financial metric—how much money a company makes—can feel like a house of mirrors. Venture capitalists say AI startups these days are experimenting with a variety of new pricing models that can paint an overly rosy financial picture.
That has frustrated venture capitalists who say they are working overtime to suss out what’s real before they make an investment decision.
Khosla Ventures Partner Ethan Choi said he evaluates 10 to 20 enterprise AI startups each week and has a 20% confidence level that their revenue is accurately portrayed, significantly below what it was before the AI boom. He said he regularly encounters deals that look promising at first glance, but once he digs into the company’s financials, he determines its revenue isn’t what it seemed.
In one instance earlier this year, Choi evaluated a buzzy San Francisco-based AI agent startup that had off-the-charts growth figures and was backed by top-tier investors. Choi pored over the company’s financials and learned it was booking short-term usage fees as annual recurring revenue, among other issues. The revelation dampened Choi’s enthusiasm and he passed on the investment.
“It is very, very tricky right now. As we get into the AI world, that quality of revenue question is what we’re trying to figure out,” Choi said. “I would take a lot of comfort in a founder that says, ‘When I talk about annual recurring revenue, let me just unpack that for you.’ They often don’t do that.”
AI has ushered in an era of innovative startup pricing strategies. Previously, startups typically stuck to traditional ARR, which had a largely uniform definition, predicated on yearly or monthly subscriptions that companies paid to use software across their enterprise. That offered investors a steady and predictable metric to analyze.
Now, AI startups have turned that model on its head.
Read the full article.
|
|
|
And now on to the news...
|
|
|
|
|
|
|
|
|
|
Raindrop is based in San Francisco. GODOFREDO A. VÁSQUEZ/ASSOCIATED PRESS
|
|
|
|
|
|
AI that monitors AI agents. In the burgeoning artificial-intelligence era, there is seemingly an AI agent for everything, from customer service to job recruiting. Now, there’s AI to help your agents. Raindrop, a young San Francisco-based startup, raised a $15 million seed funding round led by Lightspeed Venture Partners to expand its AI-powered platform that companies can use to monitor their AI agents, which are the bots that can complete tasks on behalf of users.
-
Raindrop’s AI can flag instances where a client’s agents gets tripped up and then compile reports so the company’s engineering teams can prevent future mistakes.
-
Additional investors in the round include Figma Ventures, Vercel Ventures and Y Combinator, among others.
|
|
|
|
SF Compute, an AI Computing Marketplace, Raises $40 Million
|
|
|
AI developers sell services on a usage-based pricing model, but they are often locked into rigid, long-term contracts for underlying computing resources. Many investors are worried this is a flawed model, where computing bills may end up exceeding usage and leading to bankruptcies. The startup San Francisco Compute says it has a better way: a marketplace for artificial intelligence computing capacity. The company has now collected $40 million in equity financing, valuing it at $300 million.
|
|
Trump’s Focus on Drug War Means Big Business for Defense Startups
|
|
|
The U.S. military has turned its attention southward, and the defense industry is lining up to sell it the tools for a different kind of war. Defense-tech companies and artificial-intelligence startups have found a vital new market in President Trump’s rapidly escalating drug war. Weapons and AI platforms that were designed for a future conflict with China or struggled to prove themselves on the Ukrainian battlefield have found a niche in the administration’s tech-enabled crackdown on drug trafficking.
|
|
|
|
An AI Startup Looks Toward the Post-Transformer Era
|
|
|
|
|
|
|
|
Most of the worries about an AI bubble involve investments in businesses that built their large language models and other forms of generative AI on the concept of the transformer, an innovative type of neural network that eight years ago laid the foundations for the current boom.
But behind the scenes, artificial-intelligence researchers are pushing into new approaches that could pack an even bigger payoff.
One early-stage startup developing a transformer alternative, Palo Alto, Calif.-based Pathway, plans to announce Monday that its “Baby Dragon Hatchling” architecture now runs on Nvidia AI infrastructure and Amazon Web Services’ cloud and AI tech stack.
The company doesn’t plan to actually ship Baby Dragon Hatchling until next year, but once that happens its Nvidia and AWS compatibility means companies will be able to put it into production “the next day,” Pathway said.
Baby Dragon Hatchling imbues AI with memory that large language models can’t match, according to Pathway, theoretically enabling a new class of continuously learning, adaptive AI systems.
Read the full column.
|
|
|
|
|
Correction: CB Insights said in November that it had identified over 1,700 companies, including venture-backed startups and public companies, building AI agents. A newsletter item on Nov. 18 incorrectly described all of the companies as startups.
|
|
|
|
|
|
|
|
|
|
|
People
Healthcare-focused investor Flare Capital Partners appointed Pam Walser as controller. She was most recently a finance manager at Summit Partners.
|
|
|
|
|
Harmonic, a Palo Alto, Calif.-based AI startup developing mathematical superintelligence, scored $120 million in Series C funding at a $1.45 billion post-money valuation. Investors included Sequoia Capital, Index Ventures and Kleiner Perkins.
Vijil, a Menlo Park, Calif.-based developer of trust infrastructure for AI agents, picked up a $17 million investment. BrightMind Partners led the round, which included participation from Mayfield and Gradient Ventures.
Automat, a San Francisco-headquartered enterprise workflow automation platform, secured $15.5 million in Series A financing. Felicis led the round, which saw contributions from Khosla Ventures and others.
Onepot AI, a San Francisco-based chemistry automation startup, emerged from stealth with $13 million in seed funding from investors including Khosla Ventures, Fifty Years and Speedinvest.
BHub, an accounting startup serving Brazilian entrepreneurs, added $10 million in new funding led by Next Billion Capital Partners.
Pibit.AI, a San Francisco-based insurtech startup, landed $7 million in Series A funding. Stellaris Venture Partners led the investment, which saw participation from Y Combinator and Arali Ventures.
|
|
|
|
|
|
|
|
An Anduril autonomous air vehicle on display. HOLLIE ADAMS/REUTERS
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
