In January 2026, 148,392 people went to CES in Las Vegas. Two lines in the independently audited figures published by the organizer, the Consumer Technology Association (CTA), matter more than the headline number: 39,929 attendees came for AI, up 22% year over year, and 19,605 came for robotics, up 26%. Of more than 4,100 exhibitors, about 1,200 were startups.
More people, more booths, and real signals that are harder and harder to pick out.
In January 2027, we will take up to 15 AI and hard-tech companies from around the world from CES to Silicon Valley. Why CES, why Silicon Valley as well, and why only 15?
Why the better technology does not always win
In June 1967, the first CES opened at the New York Hilton, with 117 exhibitors and 17,500 attendees. Televisions and radios were the stars. By 2027, it will have been exactly 60 years.
The story investors cite most often from those 60 years is the videotape format war. Sony's Betamax had the better picture. JVC's VHS recorded longer, had more titles and cost less, and it won the mass market. Leading on technology is not the same as winning commercially. What decides the outcome is whether product, cost, ecosystem and user needs line up.
On January 9, 2007, while CES was under way in Las Vegas, Steve Jobs unveiled the first iPhone at Macworld in San Francisco. The product that reshaped the mobile internet did not appear at CES.
CES gradually took on a different role after that. It does not produce an industry-changing product every year, but it shows a cross-section of an entire industry: where the giants are heading, what startups are trying, how far supply-chain costs have fallen, and how far a technology still is from mass adoption.
It also helps tell two kinds of products apart. The first Roku player, launched in 2008 for $99, could only play Netflix, yet it caught a growing demand for simpler, more direct home entertainment and went on to become one of the largest TV streaming platforms in North America. At the other end is Juicero, a $699 connected juicer that went on sale in 2016. In 2017, Bloomberg found that squeezing its juice packs by hand worked about as well.
One used technology to make the experience simpler. The other made a simple task complicated. Neither debuted at CES, but both directions show up on the show floor every year. For investors, telling them apart is how you find opportunities. For founders, it can decide whether the company survives.
In the AI era, demos are the one thing in oversupply
A few years ago, a robot that could walk or a pair of glasses that could recognize objects was enough to draw a crowd. Today, foundation models, chips, sensors and developer tools have matured, and building an impressive demo keeps getting easier.
Commercialization has not become any easier. These are the questions really worth answering at CES:
- What real need does a pair of AI glasses solve?
- Can a robot work reliably in a messy environment once it leaves the booth?
- Will users keep paying for an AI agent?
- When competitors can build similar features, where is the differentiation?
Investors usually look for three kinds of signals at CES:
- Supply-chain signals: when a category suddenly fills with small and mid-sized makers, key component costs have usually come down and the industry is heading into competition at scale.
- Big-company attention: when Samsung or Sony puts a technology into a core product line, the direction has been validated by the industry, and competition is about to intensify.
- Warning signs: demos that lean heavily on staging but cannot explain the user, the use case or the business model deserve caution.
These signals matter to founders too. You need to know which kind your product looks like to customers and investors. Applause at the booth only shows that someone is interested. Whether anyone will pay shows up in customers, orders and recurring revenue.
Why so many teams go to CES and come back without business
Every year, a large number of startups head to Las Vegas. They prepare booths, make demo videos, and exchange contacts with customers, journalists and investors from around the world. Four days later, they come home with a stack of photos, business cards and meeting notes.
Then a great conversation never leads to a second meeting. A prospective customer is interested but has no purchasing plan. An investor likes the technology but still does not understand the business model or the size of the market. The show is over, and the business relationships that need to move forward have only just begun.
The problem is rarely that founders are not trying hard enough. The show is too dense and too short. Opportunities that can move a business forward have to be screened in advance and followed up afterward, and a team on its first trip can hardly do that alone.
That is why, beyond CES, we put just as much weight on Silicon Valley. Las Vegas is the place to observe global technology competition in one concentrated week. Silicon Valley is home to technology companies, early customers, investors and industry decision-makers, the place to validate a market, build relationships and push partnerships forward. See the opportunity at CES, move it forward in Silicon Valley: our job is to connect the two.
From CES discovery to Silicon Valley deals
That is the starting point for the SVTR 2027 CES × Silicon Valley AI Venture Camp. It is the third cohort of the SVTR AI Venture Camp series, running January 6–12, 2027 (arrive January 5), for up to 15 AI and hard-tech startups selected worldwide.
The 7-day program has four phases:
DISCOVER | Find opportunities at CES (January 6–7) Show-floor routes are planned by sector, covering Physical AI, robotics and intelligent hardware, AI infrastructure and enterprise technology, starting from innovation areas such as Eureka Park. The focus is on competitors, supply chains, prospective customers and partners. Each day ends with a founder debrief, building a shortlist of priority opportunities.
CONNECT | Turn discoveries into business conversations (January 8–9) CES Deal Day sets up targeted 1:1 meetings with prospective customers, channels and strategic partners, followed by a curated networking dinner. On the morning of the 9th there is a final round of priority meetings, and in the afternoon the group moves to Silicon Valley.
LAND | Validate the US market in Silicon Valley (January 10–11) Workshops on US market entry, ideal customer profiles, positioning and pricing, and pitch refinement. The second day is Enterprise Buyer Day, with meetings with enterprise customers, channels and technology partners arranged around each company's sector and product. We want every founder to answer one question: if you had to find your first US customers tomorrow, where would you start?
CAPITAL | Connect with investors (January 12, Palo Alto) Investor Office Hours in the morning, the SVTR Global AI Venture Demo Day in the afternoon, and a closing dinner in the evening. Every company leaves with a 30-day follow-up plan: key contacts, business opportunities and concrete next steps.
An optional day in Napa follows on January 13, with limited places. Specific guests, host companies and meetings are subject to final confirmation. See the full 7-day program
Why only 15? Because every company's problem is different. One is looking for its first US enterprise customer, another wants to enter a new supply chain, another is preparing to raise, and another needs to rethink its product positioning for international markets. None of this can be solved with a standard tour route. Keeping the cohort small leaves room for preparation before the trip, targeted matching and follow-up afterward. Our measure for the week is simple: when founders leave, how many important things have actually started to move.
"From CES discovery to Silicon Valley deals" is the direction we work toward, not a guarantee of orders, contracts or funding.
What we learned from the first two cohorts
In March 2026, SVTR brought 15 Silicon Valley and global technology companies to AWE 2026 in Shanghai as a delegation, while more than 30 global AI founders visited six AI innovation hubs in and around Shanghai. Read the delegation recap: Fifteen Silicon Valley companies set up next to BOE.
In August 2026, 10 robotics and embodied-intelligence teams pitched and held closed-door meetings in Silicon Valley. Not one was competing in general-purpose humanoid hardware. They were building actuators, AI glasses, pet and home companion devices, a construction bidding agent and an elder-care companion platform, and several already had shipments or paying customers. Read the recap: 10 teams on stage, not one building a humanoid.
The one lesson we are surest of after two cohorts: the value of a camp is not how full the schedule is, but how relevant each conversation is to the problem the company needs to solve. A show may last only a few days, but business relationships can last much longer.
The teams we are looking for
Teams with a product, prototype or verifiable technical progress that want to advance any one of these goals:
- Market expansion: validate US customer demand and find channels, pilots or industry partners;
- Fundraising readiness: sharpen how you present the company and connect with matched investors;
- Partnership: explore product integration, technical collaboration or joint development.
Companies from China, the United States or anywhere else are welcome to apply. We recommend that the founder, CEO or a core team member attend in person.
We also do not want promising teams to miss out because of short-term budgets. That is why we created the SVTR Founder Scholarship: up to $9,000, awarded to one company, with other award amounts set through review. In addition, SVTR is an AWS Activate Provider, and eligible participating companies can apply separately through SVTR for up to $200,000 in AWS Credits, with eligibility and amounts decided by AWS.
Applications submitted by November 30 qualify for Stage 1 pricing; Stage 2 runs December 1–15, while places last. See the camp page for details.
Investors, corporate venture teams and industry partners focused on AI and hard tech are also welcome to register for this cohort's showcase and closed-door sessions, and receive the cohort's project briefs in advance (with each company's permission).
Science fiction writer William Gibson has a line that is often quoted: "The future is already here. It's just not very evenly distributed."
Every January, Las Vegas is one of the places where the future is most densely distributed. For founders, how much of the future you see matters less than whether you can bring it into a real market. In January 2027, we want to work with 15 companies to discover it at CES and close deals in Silicon Valley.
SVTR 2027 CES × Silicon Valley AI Venture Camp
- Dates: January 6–12, 2027 (arrive January 5)
- Locations: CES in Las Vegas → Silicon Valley
- Cohort: up to 15 companies selected worldwide
- Stage 1 deadline: November 30, 2026
- Application and review: free; apply for the Founder Scholarship at the same time
- Apply: svtr.ai/camp/ces-2027
- Full program: svtr.ai/camp/ces-2027/program