Jason Cui
San Francisco, California, United States
7K followers
500+ connections
7K followers
500+ connections
About
Former founder and product person. Now partner at a16z focused on investing in…
Articles by Jason
-
Thank You Uber + A New Journey
Thank You Uber + A New Journey
After almost two years, my journey at Uber has come to an end. I’m so immensely grateful for a team that has felt like…
29 Comments -
5 Ways to Utilize Your Company’s Interns
5 Ways to Utilize Your Company’s Interns
This post is on behalf of DXCP, a non-profit organization based in DC revolutionizing the way students are connected to…
6 Comments
Activity
7K followers
-
Jason Cui shared thisWelcome to team infra Christine Cheng. It's been amazing working together!!Jason Cui shared this🎺 🎺 🎺 I’m thrilled to share that I’ve joined Andreessen Horowitz as an Investing Partner on the AI Infrastructure team! AI is rebuilding the compute stack from the ground up. Models are becoming more capable, workloads are changing, and new infrastructure is emerging across compute, data, inference, developer tooling, and security. There are few moments in technology when this many foundational layers move at once, and I can’t imagine a more exciting time to become an investor. I’m also proud to have been part of the Citadel community, where the sharpest minds in finance and technology collide. The experience shaped how I think, work, and approach difficult problems, and I’m deeply grateful for the mentors, colleagues, and friends who invested in me. That experience also reinforced something I’ve always believed: the right people can transform a company. One of my favorite things has been introducing talented friends to portfolio companies I believe in and watching the right person and the right company help each other thrive. At a16z, I’ll get to do that on a much larger scale by backing exceptional founders early and helping them find the people, customers, and community they need to win. If you’re building anywhere in the AI stack, even if you’re still at the whiteboard stage, or if you’re looking to join an exceptional company, I’d love to hear from you. Let’s build 🔥
-
Jason Cui reposted thisWe’re thrilled to announce our investment in Vals. AI has gotten very good at taking tests and winning trophies. But it’s still surprisingly hard to prove that a model can actually do the work. That gap is becoming one of the most important problems in AI adoption today. Large amount of tokens and time are wasted on figuring out if we are making the right choice when choosing a model. This is because models are no longer just answering questions. They’re analyzing financial documents, conducting legal research, building software, navigating enterprise workflows, and operating autonomously for longer stretches of time. We expect them to "do the work" yet the trial period was answering through a list of multi-selection questions. The old benchmarks were the SATs and LSATs for AI. To understand real-world capability and safety, we need driving tests on the road. That’s what Vals is building: evaluations based on the work people actually want models to do. It sounds obvious. It’s technically very hard. You have to translate expert workflows into reproducible tasks, grade the final work product to the standard of someone in that profession, and run those evaluations securely and quickly enough to keep pace with frontier model launches. There’s a bigger market structure story here, too. Every market eventually needs an independent scorekeeper. When vendors grade their own homework, buyers need someone they can trust. Credit markets have ratings agencies. Public markets have auditors. Products have independent safety certification. That moment is here for AI. Every model provider can tell you where its model wins. But for an enterprise, knowing whether a model actually works for your job requires building evaluations, rerunning them as models change, catching regressions, measuring improvements, and keeping up with a relentless cadence of new releases. This is the opportunity we see in Vals: to become a trusted compass for the AI market helping every organization understand how the models are working for them, and what they can trust along the way. We’re incredibly excited to partner with Rayan K., Langston Nashold and the entire Vals team as they build the trust layer that will underpin the AI economy. More on why we invested: https://lnkd.in/eeVbMmAw And if this mission excites you, Vals is hiring: https://lnkd.in/er-R7WMXJason Cui reposted thisToday we're announcing our $40M Series A at a $400M valuation, led by Andreessen Horowitz, with participation from existing investors 8VC, Pear VC, and Bloomberg and new investors Hudson River Trading and NextLadder Ventures Alongside the fundraise, three more announcements: - Vals Smith is now generally available. Anyone can create a custom coding benchmark from any GitHub repo with 120 free credits to get started. - Frontier Risk Benchmarks: We are releasing the RSI Index in collaboration with CoreWeave and just launched ReverseEngBench, a new cyber benchmark built with Columbia University, Tufts University, UC Berkeley, and UCLA. We are also sharing our initial work in mental health, with more to come across environmental impact, military, and biosecurity. - Website + Vals Index 2.0: We completely rebuilt the Vals website and have released Vals Index 2.0, with coverage of more of the economy. The AI economy runs on self-reported grades. When a model ships, the scores come from the company that built it. No other industry this size works that way, and every side of the market is now feeling it. Labs have no credible way to prove their models are improving. Governments have no independent measure of cyber risk or the pace of US-China competition. Enterprises are ready to adopt but can't measure what they're getting back. We built Vals to be the independent evaluation layer the industry is missing. Revenue this year has already grown 8x compared to all of 2025, our customer base has doubled, and our team has tripled in 6 months. We work with all of the major foundation model labs and our benchmark results have appeared in the major model cards of OpenAI, Anthropic, Google, Meta, and xAI. Check out our new website and try Vals Smith!
-
Jason Cui shared thisVery excited to share our investment in Netris. AI infrastructure needs a new networking stack. Netris is making it dramatically easier for cloud providers and enterprises to build, automate, and operate high-performance AI networks at scale. Thrilled to partner with Alex Saroyan, Arsen Arakelyan, and Tigran Martirosya, and the entire Netris team!Jason Cui shared thisWe're excited to lead Netris' Series A. Data center networking may be changing even more quickly than compute. It is enormously complex to manage, and when it fails, a data center full of expensive GPUs stops working. Netris built a powerful and robust network management solution for GPU clouds or anyone running a large GPU data center. What impressed us most was the scale: operators run close-to-$1B data centers on Netris software - and the company has spent 8 years earning that trust. We’re excited to partner with Alex Saroyan, Arsen Arakelyan, Tigran Martirosyan, and the Netris team to take networking into the age of AI. By Guido Appenzeller, Raghu Raghuram, and Jason Cui
-
Jason Cui reposted thisJason Cui reposted thisNew market map: bringing AI assistants to iMessage 📱 The next major interface for consumer AI? Text messages. People don’t want to open an app every time they need help - they want a contact they can text like a friend. And the gold standard is iMessage. Texts sent via iMessages are much less likely to get carrier-filtered or marked as spam (vs. SMS). iMessage also gives features like typing indicators, read receipts, and link previews. And...people just like blue bubbles 😂 Historically, it's been hard for AI companies to build on iMessage. Apple has no public API. Infra providers solved this by running fleets of real Apple devices on real phone numbers. Now, Apple is opening a front door. Last week, Poke became the first AI agent approved on Apple Messages for Business 👀 More broadly - it's becoming pretty clear that we need to give AI agents / assistants anything a human in the same role would have (e.g. phone numbers, credit cards, email addresses). If you're working on something here at the app or infra layer, please reach out! Companies in this map: AI assistants on iMessage - The Interaction Company of California (Poke), Ollie AI, Nozomio, lucas ai, Orchid, tomo ai, Zo Computer, Lindy, Sauna by Wordware Infra providers - Linq, Chert, Blooio, Sendblue, AgentPhone, Photon
-
Jason Cui shared thisWe are incredibly excited to partner with Exa on this next chapter of their journey. As AI workflows proliferate exponentially, it's clear that web search has become crucial in providing agents and systems access to external info and web data. Exa is on the journey of building perfect search. The team has made incredible progress so far, and Will Bryk and Jeffrey Wang have been obsessed with search since their Harvard dorm days. We are so excited to lead their Series C! More on why we invested in the comments.Jason Cui shared thisWe’re excited to lead Exa's Series C, to back their ambition of perfecting web search, and making it ready for the age of intelligent agents. We’re entering a new era for search, built from the ground up around what AI can do. The signal from the market was consistent. Exa excels with the hardest queries – the long tail of high alpha searches where traditional engines fail. What struck us most was the default behavior: developers and agents are reaching for Exa first. Years before ChatGPT or the AI boom really started, Will Bryk and Jeffrey Wang were inspired by the transformer breakthrough and believed deeply that AI will fundamentally change the way we access information. So they set out on the path to build the search engine for a future where agents become the primary consumers of the web. We’re thrilled to partner with Will, Jeff, and the Exa team as they build the perfect search engine for agents and usher in a world of abundant information for all. By Sarah Wang, Jennifer Li, Steph Zhang, Jason Cui
-
Jason Cui shared thisExcited to moderate this panel next week at SAIR’s Science x AI summit! AI is changing not just how we build software, but how science gets done, from discovery and simulation to lab automation and data analysis. Should be a fun convo with Lan Xuezhao, Michael Stewart, Jon Chu, Tomasz Tunguz, and Ivan Zhou!Jason Cui shared thisThe capital backing the next wave of AI x Science. At the Science x AI Summit, the firms making the bets sit down for a panel on what's working, what's not, and where they're looking next. Moderated by Jason Cui, Partner at a16z, with: — Lan Xuezhao, Founder & Managing Partner at Basis Set Ventures — Michael Stewart, Managing Partner at M12, Microsoft's Venture Fund — Jon Chu, Partner at Khosla Ventures — Tomasz Tunguz, General Partner at Theory Ventures — Ivan Zhou, Partner at Accel Six of the sharpest investors in AI on one stage. The patterns they're seeing. The founders they're backing. The thesis behind every check. Register now: https://lnkd.in/gVRY5zrU #AIforScience #ScienceforAI #SAIR #VentureCapital
-
Jason Cui shared thisAI has transformed how video is created. We think the next wave is about understanding it. Over the past few years, we've seen remarkable advances in video generation, editing, avatars, and creative tooling. An increasingly important problem is teaching machines to search, analyze, reason over, and extract insight from video - across massive libraries and live streams alike. We're calling this video intelligence, and we're actively looking to back founders building here. We're most excited about companies pushing on the core capabilities: - Video-native models - multimodal embeddings, temporal reasoning, and retrieval built specifically for video rather than adapted from image or text - Real-time and large-scale pipelines - infrastructure for processing, indexing, and querying video at the speed and scale enterprises actually need - Agentic and reasoning layers - systems that don't just retrieve clips but answer questions, surface anomalies, and take action on what they see The models and infrastructure to make this real are appearing to be crossing a capability threshold right now. Multimodal foundation models are maturing, storage costs have collapsed, and enterprises are sitting on years of unstructured video with no way to use it. That infrastructure unlocks a wide range of applications including media and sports workflows, security and physical operations, enterprise knowledge management, advertising analytics, robotics, and consumer products, where video has historically been dark data. If you're building in video intelligence at the model layer, the platform layer, or in a vertical application, we'd love to talk!
-
Jason Cui shared thisIt's a pleasure to work with the SAIR Foundation as they bring together the community around AI for Science. More to come!Jason Cui shared thisJason Cui is a Partner at a16z, investing at the intersection of foundational AI, developer tooling, and scientific platforms. SAIR co-founder Chuck Ng spoke with Jason on where this moment fits in the longer arc of scientific progress — and why he believes AI for Science represents a structural shift, not a cycle. The feedback loop he describes is simple but profound: AI accelerates discovery, and deeper scientific understanding strengthens AI in return. He also shares what a16z looks for in this space — from foundation models applied to mathematics and domain sciences, to platforms enabling simulation and computational modeling at scale. And why meaningful progress here ultimately depends on collaboration across researchers, institutions, and industry. 🔗 Watch the full conversation: https://lnkd.in/gTnJhc6B #AIforScience #SAIR #a16za16z’s Jason Cui on AI for Science, Infrastructure & the Next Wave of Breakthroughsa16z’s Jason Cui on AI for Science, Infrastructure & the Next Wave of Breakthroughs
-
Jason Cui shared thisSuper excited to double down on our partnership with Braintrust. It’s been amazing watching them grow over the years, not just as a company but through all the incredible AI products their customers are building!Jason Cui shared thisBraintrust has raised an $80M Series B. We're building the infrastructure that helps teams measure, evaluate, and improve their AI products. Don't take our word for it. Hear how customers like Notion, Vercel, Navan, and BILL use Braintrust to ship quality AI. Thank you to ICONIQ for leading this round, and Andreessen Horowitz, Greylock Partners, basecase capital and Elad Gil for joining us again. https://lnkd.in/gbhuWkcu
-
Jason Cui liked thisJason Cui liked thisExa just opened our NYC office! We now have 12 amazing people in Soho, growing rapidly. I grew up in NYC, a subway ride away. So now I get to meet with customers all day and then visit my family for dinner 💙 We're finding that pretty much every NYC company needs high quality search for their AI agents: banks, hedge funds, marketing firms, recruiting shops, fashion agencies.. If you want to help redesign the way NYC finds information, DM me - we're hiring for AEs, FDEs, SDRs, and lots of other roles.
-
Jason Cui liked thisJason Cui liked thisLast week, we Andreessen Horowitz hosted a dinner with Cognition and 50 portfolio companies to talk about the future of software development, including the advantages of cloud agents, cost/performance tradeoffs across models, and how Cognition uses their agent Devin internally. Thanks to Steven Hao, Cognition's co-founder/CTO, for chatting with our portfolio, and to the Cog team who spent time with our companies after the talk! 🦦 ~90% of Cognition's code is written by Devin today, and their eng velocity has exploded while eng headcount has stayed <80, showcasing how much a lean team can accomplish with the capabilities available today. cc Chris S. Sara Xiang
-
Jason Cui liked thisIt's been an honor working with Jennifer Zhou and Ning Cao, and partner in crime Marco Mascorro, to continue pushing the frontier with better and higher quality enviroments!Jason Cui liked thisToday we're coming out of stealth with $16M in seed funding led by Andreessen Horowitz, with participation from SignalFire, South Park Commons, Scale Angels, plus angels including Fei-Fei Li, Ian Goodfellow, and Julian Schrittwieser. We're also open-sourcing Karotte, the framework behind everything we've built. Every RL environment we've shipped to frontier labs was built on Karotte, including the environments behind $15M in revenue over the last month. We've hardened it through more than a million evaluation runs and extensive red-teaming, because models are very good at finding ways to cheat. We've watched them launch fork bombs, read the answer key, and crash the grader to avoid getting scored. Karotte is designed so none of that works. When a model gets rewarded for cheating in training, it learns that cheating is a good strategy, and that lesson carries over to everything else it does. Getting environments right is one of the most important things we can do to train AI that's both capable and aligned. Jennifer and I started Preference Model because we believe data is the biggest driver of model capabilities, and that RL environments are now the most important data there is. I'm so proud of our team for building this in a year, and grateful to everyone believed in us early. Jennifer Li Marco Mascorro Martin Casado Ryan Wexler Ilya Kirnos If you want to help build the environments that keep superintelligent AI aligned, we're hiring. Links are in the comments.
-
Jason Cui liked thisJason Cui liked thisWe're hiring for four roles at Currence! - Senior Backend Engineer, AI. Build the agentic workflows and LLM-powered pipelines behind our data engine, along with the evals that make our data trustworthy. - Data Product Manager. Own the roadmap for our data engine, deciding which datasets create the most value and how customers reach them. - Account Executive to own the full sales cycle with senior buyers across energy, infrastructure, and financial services. This is for someone who wants to close deals and grow the business. - Founding Customer Success Manager: make sure customers get what they came for, from onboarding through renewal, and connect them directly with our analysts. As the first person in the role, you'll build the playbook. Come help us build market intelligence to power the AI era! Link in the first comment.
-
Jason Cui liked thisJason Cui liked thisWe’re so excited to welcome Matias Gabriel Katz Bardavid to Straia as a Staff Software Engineer! 🚀 Matias brings 20+ years of software engineering experience, with a deep background in backend systems, architecture, data, and building robust software at scale. Most recently he worked on privacy and data governance at The Trade Desk, building systems that translated complex policies into enforceable controls across the platform. As Straia continues to grow, Matias’s backend experience will be huge as we continue building real-time data infrastructure and governance layer that powers secure, trusted access to institutional data. Welcome to the team, Matias!! Excited to have you building with us 🔥
-
Jason Cui liked thisJason Cui liked thisI've been at Baseten for 4 weeks and I'm already learning so much, working with the best people, and having so much fun! I’m leading Model APIs, our hosted endpoints for the latest open-source models. We deploy and operate these models in production, serving some of the most demanding workloads for companies at the frontier of AI. The team is small, the opportunity for impact is high. If you're interested in a fast-paced startup environment and enjoy working across the stack from inference engines and runtimes through scheduling, serving, and routing, DM me!
-
Jason Cui liked thisJason Cui liked thisThat's a wrap on Tech Week Singapore. The Netris logo was on every badge at the Cloud & AI Infrastructure show, as a major sponsor alongside Nscale — and the team spent the week with AI cloud operators at booth #5-L10, discussing network automation and multi-tenancy with neoclouds, sovereign AI operators, telcos, and enterprise AI factories. We launched a customer story with Sharon AI, Inc. this week. The fireside chat video features Sharon AI, Inc. Co-Founder & Head of Strategic Partnerships Andrew Leece and Netris Co-Founder & CEO Alex Saroyan talking about how Sharon AI scales its AI cloud across multiple clusters, data centers, and network architectures — and why they standardized on Netris to deliver a consistent experience to customers regardless of the underlying architecture. We hosted "An Evening Above Singapore," an executive reception overlooking Marina Bay with our partner CloudEngine Digital. The next morning, we invited CloudEngine Digital's Quek Yi Feng 郭义丰 (Business Director) and Alan Haw (Enterprise Sales Director) to our on-site video studio for a conversation with Netris CMO Gary Sevounts. We also connected with other major partners in the region, including World Wide Technology and Hewlett Packard Enterprise (HPE), both exhibiting at the show, along with representatives from several Asia-Pacific customers, including Linton Burling, CRO at Sharon AI, Inc., and Joanna Chung-ting Cheng, Marketing Lead at Foxconn-backed Visionbay.ai. Netris Director of Technical Product Marketing Phillip Gervasi took the Keynote Theatre stage to talk about key lessons from 45+ multi-tenant GPU clouds and AI factories, including how to automate and operate the network as tenants, capacity, and hardware keep changing. Thanks to everyone who stopped by, joined our session, or shared what they're building. Next up: NVIDIA GTC Berlin, October 20–22. #TechWeekSingapore #TWS2026 #AIInfrastructure #Netris
Experience
Education
-
Harvard University
-
Activities and Societies: Harvard College Undergraduate Council, Harvard Ventures, Quad Sound Studios
-
-
-
-
-
Honors & Awards
-
Research Science Institute Scholar
Center for Excellence in Education
Selected as one of 50 U.S. high school juniors to attend the all-expenses-paid Research Science Institute summer program at MIT. Wrote a full-length research paper and presented at a symposium for distinguished faculty and research professionals from around the Boston area. Received an award for Top 5 Oral Presentations.
Other similar profiles
Explore more posts
-
Meet Global by Business Next Media
🔥 Essence Venture Capital: The Solo Fund Backed by a16z and Built Without a Pitch Deck Timothy Chen went from rejected engineer to one of Silicon Valley’s most-watched solo VCs. His latest fund? $41M. LPs? Marc Andreessen, Martin Casado, and Bain. What changed? 💻 From Microsoft coder to startup founder (exit to Cloudera) 🚀 Angel investor turned solo GP after Sand Hill said “too technical” 🎯 Now backs early-stage infra/devtools startups—often before product 📄 Full story in the comments. #VentureCapital #EssenceVC #Startups #AI #DeveloperTools #Infra
2 Comments -
Christopher McKelvy
HHS just launched an AI prize competition for caregiving startups. What they’re looking for: - give caregivers on-demand support/training - help monitor well-being safely - cut admin + workflow automation Phase 1 is up to $2.5M, applications are due July 31 and they expect to give 20 awards. Caregiving is one of the biggest pressure points in America. Immigration policies have strained the workforce yet demand is growing (we have a lot of Baby Boomers). AI isn’t a silver bullet but, given the size of this market, any automation will help. If you’re building in aging, disability care, or home care, this is worth a look.
16 Comments -
Olivia Moore
AI is here to transform marketplaces. But it's not creating new ideas - it's taking old ones, and finally making them work. Marketplace categories that were previously graveyards are now greenfields, with space to build massive companies. New piece from our team Andreessen Horowitz - featuring Jack & Jill, Dex, Spotlight Realty (YC S25), Honey Homes, Remi, Lawhive, Paraform, Counsel Health and more: https://lnkd.in/gBE4tDTN
15 Comments -
Mohammad Sudaise shah
🤖 This VC’s best advice for building a founding team One of the most consequential decisions early-stage founders have to make is who they will bring on as their founding team. The first five to 10 employees will have a massive impact on the company culture, and the precedents set with them are difficult to change down the road. That’s why this season on Build […]
-
Dave Messina
I've looked at a lot of YC batches. The bio and health lineup in P26 is STACKED. https://lnkd.in/gQKxpe9t Adialante - Compact, mobile whole-body MRI to make cancer screening accessible at hundreds per scan FinalDose - A programmable drug that reads DNA inside the cell and destroys it if it's diseased, starting with previously undruggable cancer targets Voquill (YC P26) - An AI coworker for pathologists that learns your reporting style and produces sign-out ready reports in real time Clara - An AI primary care doctor with a licensed clinician reviewing every medical decision Lumius (YC P26) - Affordable, real-time 3D ultrasound, starting with vascular access
7 Comments
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content