Anker AI Hackathon: From real problems to working solutions in 24 hours, 33 teams compete for 4 championships

Today, these scenarios are by no means empty talk—they are outstanding works produced by contestants in the 24-hour Anker AI Hackathon. These works have mature solution concepts and are even based on real Anker smart hardware products, and they truly "ran" on-site at the competition.

Behind the scenarios and works is a major competition: more than 1,800 registrations from over 180 universities and developer communities went through rounds of selection, and ultimately 33 teams made it to the offline finals, completing the transformation from ideas to demonstrable, verifiable results within 24 hours, with 4 champion teams emerging.

In the past, producing a runnable version like this in 24 hours was itself the biggest highlight of a hackathon. But today, with AI assistance, the cost of "quickly writing a demo" is approaching zero. The Stack Overflow 2025 Developer Survey shows that 84% of developers are using or planning to use AI tools. What is truly scarce at hackathons is no longer coding ability, but real problems worth solving, interfaces that can connect to the real world, and where the works go after they are finished.

Anker's first hackathon provides a new observable sample.

1. The value of 24 hours is being redefined

"The internet changed distance; the AI era changes time." This is Anker Innovations CIO Neil's judgment on AI.

Anker itself verified this statement first. Anker Innovations' 2026 semi-annual report shows that over the past year the company's total token consumption reached 30 trillion, with employees consuming an average of 450 billion tokens per day, a 30-fold year-on-year increase; AI handled more than 95% of basic code generation, saving about 80% of development work hours.

In the Workshop sharing at the hackathon, we saw that makers can use the Launch platform to quickly put roadshow works online, and works can be directly converted into accessible links. From 360-degree hardware teardown displays, product usage simulation animations, and pixel-style interactive demos to generating real-time updated BI dashboards connected to dynamic data sources, the platform has every function imaginable.

Such excellent products stem from the rapid iteration of Anker's internal "AI infrastructure": based on AI-assisted development, in mid-2025 the team completed the first prototype in just a few days; during this year's Qingming Festival holiday, the team completed full-feature launch in only 3 days and officially promoted it company-wide. From several days to 3 days, what was compressed is exactly the "time" Neil spoke of.

Within the 24-hour hackathon format, this means: after AI compresses coding time, what sets contestants apart is no longer who types faster, but things beyond code—whether the idea is feasible, whether the usage scenario is chosen correctly, and whether user pain points and problems are grasped accurately.

Anker Hackathon opened real scenarios, real problems, and real product SDKs to contestants, giving many excellent ideas and works real soil to grow in. Behind this is Anker's long-term firm investment in AI and its active pursuit of extreme innovation.

So what exactly do the "real problems" and "real interfaces" that truly set people apart beyond code look like? Let us focus on the champion works of this hackathon.

2. "Deconstructing" the finalist works: how does the ideal translate into reality?

To understand why these champion works won, we first need a different yardstick. In the view of AJ, founder of WaytoAGI, models determine the upper limit of AI, while scenarios determine the value of AI. In other words, coding ability is now available to everyone; what truly sets people apart is whether contestants have accurately identified "real problems worth solving" and whether they have put them into "solutions that can run through in the real world."

▲AJ, founder of WaytoAGI

The scoring criteria of this competition also confirm this. Product innovation and AI nativeness account for 35%, and user value and scenario implementation also account for 35%, together nearly 70%. What ultimately stands out and is memorable is precisely the works that make the word "real" real. They can innovatively solve real demand pain points and have implementation value.

During the two-day event, we were on-site and saw makers engaged in intense discussion, collision of inspiration, and all-night development. In the final roadshow, many impressive works emerged across the four tracks: smart recording, smart security, charging and energy storage, and smart services.

During actual life and study, team members found that many students are ashamed to raise their hands to ask questions for fear of disturbing others, which affects students' digestion and absorption of classroom content.

To this end, they used AI technology and smart recording devices to provide targeted AI explanations and Q&A, improving learning efficiency and teaching quality. At the same time, this product can also help teachers better understand students' learning status and obtain timely feedback and analysis of teaching effects, further assisting in optimizing teaching strategies.

The relevant person in charge of Anker's recording bean team among the judges specifically mentioned in the on-site comment session that the pain point solved by this work is highly consistent with the functions in their actual R&D. This means the direction of the work is indeed feasible and is expected to be implemented and verified in the future.

In the smart security field, where Anker has long been deeply engaged in overseas markets, the "Station Tracing" team set its sights on the thorny logistics problem of "lost packages."

The team used Anker eufy smart cameras, open product SDKs, and AI technology to conduct intelligent analysis of multiple video streams, achieving efficient identification and tracking of lost packages and solving the pain points of inefficient and inaccurate manual review in traditional logistics monitoring.

Team members specifically mentioned that after testing, they found Anker's cameras to be easy to install and low in power consumption. In practical applications, Anker devices' dynamic tracking of people and time-keyword extraction functions best met their needs. The team leveraged Anker products' existing capabilities to simplify the entire product implementation process and solved the challenges of low power consumption and multi-channel concurrency.

As mentioned earlier, Anker's massive product lineup can gather a large amount of real feedback from actual consumer use. This feedback can provide a real data foundation for building intelligent service products.

The champion of the intelligent service track, the OntoCare Ontology Team, built a service ontology and a decision ontology, adding business rules and SKUs as graph nodes to achieve rapid SOP updates and iterations. Compared with traditional Workflow and Prompt solutions, this approach can better avoid underlying code refactoring issues when facing changes in policies and regulations across multiple countries.

In addition, through red-team/blue-team exercises and case-driven development, the team developed a self-evolving system that raised the customer service issue resolution rate from 83% to 98%.

In post-competition interviews, we learned that the inspiration for this project came from the challenges faced in Anker's actual customer service scenarios. The team identified pain points from the real customer service system and validated them with real after-sales data.

In these outstanding team cases, we can see that, based on AI technology and Anker's underlying support such as intelligent hardware and SDKs, Builders can quickly turn their ideas into reality. More importantly, these ideas originate from real problems in real scenarios and have the opportunity to evolve into technologies that tens of millions of users can actually use.

In today's AI entrepreneurship wave, the biggest trap is often "looking for a nail with a hammer": starting with AI technology and then forcing it onto application scenarios. The outstanding works at Anker's hackathon, however, are based on Anker's hundreds of millions of devices. Once these applications are implemented, they can form a data flywheel in real scenarios, further accelerating the iteration of AI technology. This is also where Anker's enormous potential in the AI era lies.

First, from the competition topics, scenarios, products, and technologies to the final evaluation criteria, everything is closely centered on the word "real."

The topics of the four tracks all stem from Anker's real product lines and real business needs. And fields such as smart recording, smart security, charging and energy storage, and intelligent customer service almost all share a common characteristic: they are by no means tracks that can be done well in a short time. They all require long-term substantial investment, solid product deployment, real market share, and users.

These are not paper competition topics, but real products that Anker Innovations has built over more than a decade, that tens of millions of households worldwide are using, and that have already entered the market. They have a source and a foundation.

Second, another rare quality is the confidence to open up its "family assets" to participants.

Anker dares to open the interfaces that are currently serving hundreds of millions of users worldwide to the participants of this hackathon, demonstrating the company's strong technical confidence.

For Builders, these SDKs are truly useful, easy-to-use, and practical SDKs. These key capabilities are quickly invoked through the SDKs, enabling many ideas to be realized on real products. This would be impossible without the extremely high completeness and security of Anker's underlying technical interfaces.

Participants do not start from scratch; instead, they innovate on top of a mature technology and product foundation. Across the industry, Anker's courage to open its foundation in a competition will undoubtedly help the company better attract global innovative talent.

Third, another rare quality is the extreme discovery of outstanding talent: no thresholds, with a top-tier judging panel including the company's top leader engaging directly with makers.

It is understood that this AI hackathon uses "registration equals review," with no thresholds for educational background, background, or major, looking only at innovation ability and hands-on practice ability. Recruitment covered more than 180 universities and over 40 developer communities.

From Anker Innovations founder Yang Meng, Geekbang Technology founder and CEO Huo Taiwen, WaytoAGI founder AJ, and Linear Capital founder and CEO Wang Huai, to Amazon Web Services China Industry Cluster General Manager Tian Feng, from smart hardware giants and cloud giants to AI developer communities and investment institutions, the makers at the AI hackathon can have the opportunity to engage in direct dialogue with a top-tier industry judging lineup.

Fourth, outstanding participants and works all "have somewhere to go."

Admittedly, the total prize pool of more than 180,000 yuan and the 30,000 yuan prize for each track champion are highly motivating for makers. Outstanding contestants have the opportunity to go directly to the final interview and obtain a retained SSP Offer opportunity; shortlisted contestants will also preferentially enter Anker's high-potential talent pool.

Compared with most simulated competitions' paper-talk-style "idea shows," what the industry really needs is real soil that can transform "castles-in-the-air AI ideas" into "products that can run in the physical world." Anker's hackathon allows ideas to move from phased results toward longer-term implementation exploration, with long-term extension value.

As Huang Pengfei, Director of Automotive and Manufacturing Industry Architects at Amazon Web Services, said in his opening speech, Anker's DNA is to "take one category to the extreme, then extend the capability to the next category," and this spirit of "being meticulous" runs throughout.

▲Huang Pengfei, Director of Automotive and Manufacturing Industry Architects at Amazon Web Services

Looking at the industry's development, this hackathon is more like a "microcosm" of the current evolution of the AI industry: under the rapid advance of large models, blindly competing on parameters, benchmark scores, and rankings has become a thing of the past. The industry's core focus is gradually shifting to whether AI can be implemented in the real world and achieve commercial value.

From a recording bean clipped to a collar and a smart camera stationed at a courier site to a smart energy storage system running efficiently in European homes, after 24 hours of intensive development and impressive demos, these outstanding projects really do have a chance to enter real product development. Real hardware is becoming AI's "senses" and "limbs," and AI is no longer a feature bolted onto hardware, but a solution grown out of a specific scenario.

For individual builders, this means the coordinates of capability are shifting. "Knowing how to write code" is no longer where the gap is opened. From this year's champion projects, a common path can be seen: start from concrete problems around you, put the solution to the test on real data and devices, and reuse existing capabilities as much as possible rather than rebuilding from scratch. When the time spent writing code is compressed by AI, writing fast is no longer an advantage—judging accurately is.

For the hardware industry, this may also be one answer to building hardware in the AI era. When models can be rented and code can be generated, the value of hardware no longer depends only on specs and category, but increasingly on how many real scenarios it can bring AI into and how many real problems it can solve. Where the device is, the scenario is, and that is where the value uplift brought by AI lies.

In an excellent hackathon, the end of the competition is by no means the finish line, but the starting point from which countless ideas take flight. A hackathon is not a PR-style marketing gimmick, but a "touchstone" for testing the hardness of a tech company's AI technology foundation. Anker's attempt this time may not be the only answer, but it at least shows that innovation in the AI era must ultimately be tested in real scenarios.