Woo hoo - I’m so pumped that we finally have tech companies filing S-1s to go public again! While both Instacart and Klaviyo are not enterprise infra cos, Klaviyo is a SaaS marketing automation platform that will be a nice gauge for other enterprise companies contemplating IPOs. To illustrate how far we’ve come from the prior batch of IPOs, both Klaviyo and Instacart are profitable. Let’s see how Wall Street reacts, and hopefully it will kickstart a whole new class of efficient growth companies. I won’t break down the numbers here but Jamin Ball highlights a few key stats for Klaviyo: $585M LTM Revenue (+57%), 51% last Q YoY growth, 75% gross margin, 8% FCF Margin, 119% net retention. These are super impressive to say the least.

Speaking of the public markets, I’ve been at this for 27 years, and never have I seen a beat and raise like this from a company at this scale - Nvidia - absolutely 🤯📈.

We just have to remember that cloud companies buying shit tons of GPUs does not necessarily translate to enterprise software revenue…yet. AI costs a ton of 💰 and folks are still trying to figure out the right business model for this. That being said, it’s encouraging that even Nvidia is not just relying on the cloud providers to spur growth as it’s been heavily investing in the enterprise. This will take time to filter through but laying down the foundation to deliver models to the data versus the other way around positions Nvidia well to capture “accelerated compute” anywhere.

Partnerships with companies like Snowflake, Microsoft, ServiceNow, Accenture and now VMware will put Nvidia in the cloud and also on-prem as enterprises start spinning up new generative AI applications. The good news is that AI continues to be top of mind for CIOs and the requisite spending on enterprise software will follow this insane spend on GPUs. According to a recent survey of 147 CIOs from Piper Sandler, “generative AI rose nine spots in priority to become the top emerging technology trend for the next three years with 75% of CIOs either testing or implementing projects. Enterprise testing and implementation of Gen AI appears to be consuming incremental budget dollars,” Piper’s team said.”

This ties well with the most recently released Gartner Hype Cycle for Emerging Technologies with Generative AI at the Peak of Inflated Expectations which means that it will still take 2-5 years away for many of these projects to go from pilot to production.

IMO, one of the huge unlocks to go from pilot to production in the enterprise will be solving AI security issues. To that end, if you’re interested in learning more about this topic, I suggest reading the latest edition of the from Tamar Yehoshua (IVP).

Practical Intelligence

What Does ‘AI in Security’ Actually Mean?

For leaders in the security space, there’s a lot of pressure to incorporate generative AI into security systems. If we don’t use AI, the thinking goes, the attackers eventually will. Large language models are being used to increase productivity in many domains, and, unfortunately, they can also be used by bad actors to generate increasingly sophisticated…

3 years ago · 2 likes · Tamar Yehoshua

As always, 🙏🏼 for reading and please share with your friends and colleagues.

Share What's Hot in Enterprise IT/VC

Scaling Startups

To YC or not to YC
❤️ this from Tim Ferriss - founders who are building something big need to be a bit crazy - if everyone answers yes, I must have this, then perhaps you are only building something incremental? Having some controversy is good and great founders learn to hear but ignore the haters…
I’m going to propose that you develop an uncanny ability to be selectively ignorant. Ignorance may be bliss, but it is also practical. It is imperative that you learn to ignore or redirect all information and interruptions that are irrelevant, unimportant, or unactionable. Most are all three.
Lifestyle design is based on massive action—output. Increased output necessitates decreased input.
Most information is time-consuming, negative, irrelevant to your goals, and outside of your influence.
Nine sources of advantage from Farnam Street newsletter:
1. Raw talent/intelligence – Some people are just naturally better and smarter.
2. Hard work – Some people work harder.
3. Differentiation – Seeing the world differently. Doing something different. Reading different books. Interpreting the same information differently.
4. Process / Discipline – Creating a process and following it. Working out every day is a great example.
5. Talent Collector – The ability to hire the best people and get the most out of them.
6. Patience – A lack of patience changes the outcome.
7. Ability to take pain – Are you willing to look like an idiot to get better? How much risk are you willing to take, AND, importantly, can you handle the losses?
8. Temperament – Keeping your head when everyone else is losing theirs.
9. Luck
Most of these are within your control.

Enterprise Tech

🤯 Hugging Face, Github for ML models, just raised at 100x forward ARR multiple led by Salesforce Ventures (The Information) with a consortium of strategic investors Google, Amazon, Nvidia, Salesforce, AMD, Intel, IBM and Qualcomm.
Salesforce is paying a high price for a piece of Hugging Face, which runs a service that helps companies store and use AI software, similar to the way GitHub lets developers store software code. The new funding valued the startup at more than 100 times its annualized revenue, a measure of how much revenue the company would generate over the next 12 months at its current rate, one of the people said.
Hugging Face, for its part, is on pace to generate more than $30 million in revenue annually, one of the people said. Amazon Web Services, Microsoft, IBM and others pay Hugging Face for steering its users to their respective cloud computing services. Founded in 2016, Hugging Face also charges developers for an enterprise version of its repository for machine-learning models. The startup, which has more than 200 employees, says more than 10,000 companies use its free or paid repositories. It hosts at least several hundred thousand AI models, including popular open-source large-language models such as Meta Platforms’ Llama 2. Sequoia Capital, Lux Capital and Coatue Management last valued the company at around $2 billion in a financing last year.
So many startups have been built off of what Google/Facebook have built internally for developer tooling but why no developer cloud (Scott Kennedy - ex-Googler, h/t Gergely Orosz) and what’s needed to build the right one?
So we’ve established that the idea isn’t new. We’ve also seen dozens of companies successfully take Google/Facebook’s internal tooling ideas and build billion dollar B2B SaaS companies⁶. So where is my developer cloud?
It doesn’t exist because it’s really fucking hard.
Google can’t make theirs available because they can’t disentangle from their internal architecture without damaging their internal productivity. I know because I was there when they tried. It’s hard.
Startups or open source historically focused on one piece of the puzzle.
Sourcegraph has been working hard for a decade just to get the CodeSearch piece right⁷. Bazel might be open source, but it has a “batteries not included” feel.
The full experience requires putting many pieces together. It’s hard.
But I really want the developer cloud to exist. And I want to help build it. That’s why I chose Replit.
Fine tuning is here for GPT 3.5 - data matters, huge news but…
VC is a Power Law business which means only a few cos deliver most of the returns - David Clark from VenCap breaks it down with data on 11.350 companies from 259 funds from 1986 - 2018. Only 1.1% out of 11,350 cos returned the whole fund!

David Clark responded to my tweet that median fund size for fund returners was $350M and many funds >$1b also had fund returners or 🐲. Simple math, of course, says easier to have a fund returner the smaller the fund you have and so much harder as you scale.
On AI agents or not…so agree with Zach - brittle, riddled with errors, security issues…good for prototyping
Meta just released a specialized version of Llama called Code Llama fine tuned for generating and discussing code.
How enterprises may deploy AI models On-prem with a privacy first component as envisioned by VMware - to be released in 2024…idea is to deliver AI apps to the data vs. the other way around

Markets

❤️ how Eric, founder of Ramp, addressed his 30% down round head on (from $8.1B to $5.5B) - as I’ve said numerous times, this is just the beginning and 👏🏼 choice to do a clean round to continue building + reset valuation for new hires…
There’s an old VC adage: "Give me a valuation & I’ll give you the terms" This round was straightforward. No ratchets. No participating preferred. No kickbacks that artificially inflate valuation or misalign incentives of investors, management, employees & customers. Preferred and common were priced equally.
Rumor mill - Wiz looking at buying Sentinel One

What’s Hot 🔥 in AI/Infra/VC

What's 🔥 in Enterprise IT/VC #356