Last week, DTC hosted our first-ever AI Infra Innovation Summit along with our friends at Goldman Sachs. We brought together 60+ founders and operators building the layers underneath every AI headline.
We heard: AI is no longer a model story. It's a systems story, a data story, and a process story.
Some takeaways:
🔹 Silicon is a systems problem. Interconnect, memory, packaging, and optics all have to work together, which is why hardware is cool again. Plan supply two to three years out, and don't mistake scarcity for advantage.
🔹 Welcome to Data 2.0. The work has moved from high-volume labeling to expert-level post-training and RL. As models commoditize, protecting your proprietary data is how you differentiate.
🔹 Enterprise adoption is just getting started. AI is about process, nothing more, nothing less. The enterprises seeing returns tie every project to a measurable outcome and fix their data first. Buyers expect production in weeks, not years, which is great news for startups. Founders: build horizontal.
🔹 The M&A window is open. Buy is replacing build, and the buyer universe now reaches well beyond big tech. Things move fast. Keep your exit strategy top of mind.
Thank you to our speakers John Roese, Adam Benjamin, Marc Boroditsky, George Cameron, and Alex Ratner for their candor and insight, and to our co-hosts at Goldman Sachs, Benny Pinkas, Matt Margolin, and Karthik Subramanian, for a fantastic partnership. And thanks to everyone who joined the conversation.
Full recap here: https://t.co/TVXkZasxHb
#AIInfrastructure #AI #Silicon #Data #EnterpriseAI #Neocloud #VentureCapital
The word "thrilled" tends to get thrown around a lot in VC. Thrilled to invest, thrilled to announce, thrilled to celebrate an exit. In this case, there really is no other word to use. We *are* thrilled to welcome Jordan Garside as Senior Counsel and Head of Legal for DTC! 🎉
Jordan leads legal for DTC. His path to venture started in law school at UVA where started angel investing in sports tech alongside a Wake Forest teammate. What he loves most about venture is the founders: “There’s nothing I enjoy more than helping them get where they’re trying to go.”
Welcome, Jordan!
Hyperbolic through the lens of math.
Our CEO @zjasper on where our name comes from, what inference taught us about utilization, and why our strategy comes down to making compute unified, productive, and liquid.
Imagine a world where innovators could spend more time capitalizing on all the GPU capacity that's actually available versus scrambling to find it. That's what @hyperbolic_labs is creating with their Forge platform, which standardizes machines/cycles from heterogeneous providers, and Flux, their new inference product, that runs on top.
We're proud be a part of the company's latest funding round and look forward to where @zjasper and team take Hyperbolic from here!
We’ve raised an oversubscribed $30M Series A1 led by @walden_catalyst to build the Global Compute Grid.
We’re turning fragmented GPU capacity into one unified platform where AI teams can access, manage, sell, and orchestrate compute.
Our mission: make the world’s compute unified, productive and liquid.
🫧 🏆 🗞️ Happy to share that @TwineSecurity has been named a @ciotimes Leading Cybersecurity!
"The work gets done continuously and business value becomes something you can state plainly: risk retired per dollar".
Twine's Co-Founder and CEO Benny Porat was interviewed for the cover story, 'Closing the AI Execution Gap in Cybersecurity', which you can read online here: https://t.co/J00xEEDOxV or flip through the magazine edition here: https://t.co/6j3o038rU5
Mave is betting its AI agents can protect companies from increasingly sophisticated cyberattacks. Business Insider got an exclusive look at the pitch deck it used to raise funding. https://t.co/LIDvTQlHA6
Arena has raised a $200M Series B at a $3.1B valuation.
We are using this momentum to evaluate agentic utility in real workflows from every economically valuable industry on Earth.
We seek to give everyone universal and frictionless access to the best intelligence. The @arena platform has 10s of millions of monthly users doing complex coding and agentic work tasks. We have gone way beyond human preference, and most of our tokens go to deep, single threaded workflows with dozens of turns where we can evaluate agents while humans are in their true flow state. And we make leaderboards by observing whether they get their jobs done. This is how benchmarks should be: representative of actual human value, so we incentivize AI to benefit humanity.
Today, we are releasing one of the most exciting developments in the history of our platform. The Arena Alignment Index. In observing how humans collaborate with agents, we have witnessed that in a surprisingly large fraction of cases, agents are attempting to deceive users or take unauthorized actions beyond what the user intended or permissioned. Agents are deleting files and lying to users about what they did or didn’t do. The longer conversations go, the higher the probability that an agent will violate alignment. The results are really striking; see the plot below. And because this represents in-the-wild usage, the probability that those of us reading this have experienced some form of misalignment from our agents is nearly 100%. And we may not even have noticed.
This is exactly why it’s important to have a neutral 3rd party that is evaluating AI alignment on the distribution of actual user-agent interactions. Arena is excited to step into that role.
We are also incredibly grateful for the incredible and experienced investors backing our vision. Our Series B is co-led by @lightspeedvp and @khoslaventures with participation from @SalesforceVC, @01Advisors, @DellTechCapital and @endvr_catalyst. This round includes support from our existing investors @a16z, @felicis, @amppublic, @QuantumLightVC, @thehousefund, and others.
Which brings me to the real headline: we're hiring, across ML, product engineering, design, marketing, BD, and more. If evaluation and alignment at the frontier is your thing, and you’re looking for an environment with exceptional research and engineering, come join us. Job board link in @arena’s bio and in the thread below
LFG (safely)!
SF Tech Week schedule in one word: intense.
Not to add fuel to the fire but we've got several events you're going to want to RSVP for.
First up, Robots & Hardware at the @Tesla Showroom co-hosted by us, @Mondaydotcom, and the AI Infra builders from Tesla.
https://t.co/66OD093jHJ
Get there early if you want in on the morning pastries. Stay on for smart talk on AI Infra.
As AI adoption accelerates, enterprise leaders are under pressure to understand which initiatives are creating business value and how to optimize both investment and adoption.
But if you ask around, teasing ROI metrics out of AI deployments to date hasn't been easy.
That's why we led Ascerta's $18 million Series A. David Tepper was managing AI tech spend at one of the biggest tech companies in the world when ChatGPT hit. He realized immediately that the management of AI engagements was going to be a challenge to every enterprise at scale.
With Ascerta, David and co-founders Erik Winters and Doron Holan, and their team are building the system of record for AI value creation, helping organizations connect AI usage and activity to measurable business outcomes.
We are glad to be a part of this team's journey. Welcome, Ascerta! https://t.co/7khhfExH29
Today we welcome Shane Earley to DTC's Portfolio Development Team. Shane brings more than a decade of experience in enterprise sales, strategy and partnerships to the role. He'll work alongside Chris Falloon and Julia Beech to support GTM efforts for our portfolio companies including PMF analysis, sales leadership coaching, and direct customer introductions. We're excited to have him with us!
While most people are focused on model innovation, enterprises are focused on outcomes.
At SlatorCon San Francisco, DTC Partner Radhika Malik joined Esther Bond and Kanu Gulati to discuss where value is emerging across the Voice AI stack.
What bubbled up in the conversation:
➡️ The model layer is critical, but lasting value is increasingly created through infrastructure, workflows, integrations, and the systems that help enterprises deploy AI effectively. As Radhika put it: "The model layer is necessary but not sufficient."
➡️Enterprise adoption, not model performance alone, will define the next phase of AI.
➡️ Faster development cycles are changing how investors evaluate startups. A polished demo is no longer enough.
Thanks to @slatornews for the opportunity and for capturing the discussion in this recap.
https://t.co/vK2TIN3UMB
Congratulations to Aitan and @terraoffsec for being named to the 25 Most Promising Startups 2026 list by @N12News, @ValleyBank, and @Leumitech. We couldn't agree more!
Anyone who's spent hours trying to debug why an AI agent failed will appreciate what @OpenObserve released today. If that resonates, give them a look and show them some support on @ProductHunt today.
https://t.co/q2ikrjxlTr
The strong analyst and ecosystem support surrounding @LightbitsLabs' announcement today of Inferra highlights a growing industry consensus: inference economics matter. A lot.
Maximizing GPU utilization, reducing latency, and overcoming memory bottlenecks will play a significant role in determining how AI continues to scale across the enterprise.
We are excited to see the Lightbits team continue innovating in this important layer of the AI infrastructure stack. Congratulations to Avigdor Willenz, Eran Kirzner, and the entire team on the introduction of Inferra.
https://t.co/PxO0iE7kPf
With investments in Bland, @Cartesia, and LILT AI, you could say we're a bit bullish on the future of Voice AI.
Catch DTC's Radhika Malik and @kanugulati of @khoslaventures with Esther Bond at SlatorCon at September 3rd on the leading edge(s) of voice AI, what's being developed, where capital is being allocated, and how research and capital continue to fuel SF's AI leadership.
Enterprise AI projects often stumble because the context of how work gets done isn't there. That's the premise behind Skan's approach to AI transformation. Their work is resonating with some of the biggest banks in the US and a quarter of the Fortune 50.
This week, we re-upped our support of the team by co-leading the company's Series C announced.
In our latest interview, Skan co-founders Avinash Misra and Manish Garg share their perspective on:
✅ The biggest mistakes enterprises make when deploying agents
✅ What separates successful AI transformations from failed pilots
✅ How organizations can identify where AI delivers real ROI
✅ Why the future is human + agent collaboration, not human vs. agent
Read the full interview in the first comment.
Building a great product is only half the battle. Turning that product into a repeatable, scalable revenue engine is where many enterprise startups struggle.
That's why DTC invests heavily in portfolio development. Every year, our PortDev team facilitates hundreds of customer introductions, hosts dozens of customer-facing events, and provides hundreds of hours of hands-on go-to-market coaching to help founders navigate enterprise sales and accelerate growth.
The PortDev team was founded and built out by DTC's Chris Hillock and is now led by Chris Falloon. We talked with Falloon about what challenges founders are facing today and his plans for the development function for DTC. Read on: https://t.co/2WjkgpmnIx
Headed to Black Hat next week? Avoid the Vegas heat and join us, fellow cybersecurity practitioners, along with folks from Endor Labs, Terra Security, and The Redesign Group for a rare wine tasting event. Drop us a note for an invite.
👉 https://t.co/8TSoCtKTNp
When you're a startup, one of the most effective ways to cut through the noise is to have customers willing to share their success stories with others.
Flexor is firing on all cylinders on this and had not one but two customers go on the record this week with how the company's AI Context Engine (ACE) is playing an integral role in their AI strategies.
MLB team Cincinnati @Reds picked Flexor ACE to layer on top of @Databricks to automatically process massive amounts of unstructured player and game data to get faster, game-changing insights.
Then in a fast follow, @Scotiabank's Yashaswi V. joined Flexor CEO Or Zabludowski on stage at CDOIQ Program to talk about making unstructured enterprise data ready to work across teams at multinational bank.
Attending CDOIQ? You can still catch the Flexor team there today and tomorrow.
https://t.co/cqqFQUwDuz
DTC's Daniel Docter spent time with Mary Ann Azevedo for @crunchbasenews to talk what's new with the AI-driven innovation wave versus the many waves that have come before, DTC's quiet run of exits*, and the one piece of advice he gives every founder.
Read on for the full interview:
https://t.co/Ulw54Kn5la
*Netskope, SingleStore, Rivos, LayerX, Entro Security
cc: @bayareawriter