Why I’m So Excited About KanKan and Remark Holdings…

Tailwinds’ Guest Post…the blog piece below was submitted by a current MARK shareholder who would like to remain anonymous. This author has extensive experience in software and has developed a very deep understanding of Remark and their technology.

While I do not think too much of myself it is perhaps noteworthy to share that the only career I have had for the last 36 years is enterprise software. The first 29 years of which at a big public global software company where I enjoyed several global leadership roles. Since then I have been dedicated to my natural instincts as a software entrepreneur (watch this space).

Put frankly, I have never been as excited about any public investment in my career as that of Remark Holdings (MARK), due to its AI offering KanKan. KanKan has been years in the making and is a vision of the CEO. Translating that vision into a reality where the software architecture is world class & truly scalable in the AI space is some achievement. Couple that with the design inherently leveraging KanKan’s unique and massive active data, we have in front of us, what I believe to be, genius work that is most relevant for almost unlimited use cases of horizontal AI. Furthermore, it is important to note that one of the several benefits of this world class architecture and design is that unique, scalable & relevant KanKan Products with significant TAMs (Total Addressable Markets) can be built quickly. Really quickly. In weeks in many cases. Currently, as best as I can determine, there are 13 unique KanKan products.

As a result of KanKan, Remark has entered into unique partnerships with Alibaba, Tencent, Weibo & ACXIOM – there may be more that are not public yet or indeed I may have missed some. It is indicative of the appeal of KanKan, by way of example, that Alibaba shares its transactional data with Remark for KanKan.

There have been several KanKan recent announcements that are eye-catching.  A list is included at the bottom of this note. 

There are some specific competitors in clear subsets / submarket of KanKan. For example Face++ is specialized in facial recognition only and is strong in applying its high precision facial detection and recognition in various applications. Note Face++ recently raised money at a $2Bn valuation. Link attached below. 

It is useful to have such a reference point as this provides a platform (pardon the pun) to begin to articulate the increased opportunities KanKan provides through a deeper understanding of the KanKan AI platform. Some of the reasons why KanKan is unique, if you like. 

KanKan is able to offer the same kind of applications as Face++ with similar precisions, however, KanKan’s abilities are much more broader:

  1. KanKan Computer vision AI focuses on object recognition, whether it is a face, human/human gesture, animal, uniform, wearables, cars, crowds, plants etc with high precision, which enable us to apply this technology in much broader industries, not just safety/security, but also medical, CRM, FinTech, Employee management, inventory management, education etc.
  2. KanKan Computer vision AI offers customizable software and hardware systems that can be used to enable the existing devices (like camera, digital machinery, POS, mobile phones, drones etc.) to have AI applications right away without replacing the existing devices.
  3. KanKan AI platform not only provide realtime recognition capabilities similar or better than our competitors, but also provide data collecting, processing, analyzing, and storage capabilities for businesses require to manage their existing or future assets (movies, videos, speeches, security logs, operating data, device data) along with our AI ’s auto-recognition, auto-tagging, indexing capabilities. For example, scanning through all the video footages and auto tag the artists, cars, animals, locations, landmarks, costumes, speeches, musics and other objects in them that can be recognized by KanKan AI, so that those video footage assets can be easily found, managed, reused, shared or sold.
  4. Since KanKan AI was trained via billions of social data, KanKan AI naturally offers NLP and other important AI components to enhance the solutions that customers may want in just one package.
  5. KanKan has a proprietary  migration training mechanism/algorithms, which enables KanKan to train customized models in a very short turn around period (even within 2 weeks) with much smaller sets of samples  (instead of /comparing to  training the customized models  from scratch with millions of samples picked by human, which will take months even years) 
  6. KanKan AI is backed by KanKan Data platform, which has not only the social, O2O, travel, public government, company data that we collected for training, but also with the models trained with partners like Ali, Tencent, Axciom and others.

The opportunities that Remark Holdings has ahead of itself as a result of its unique Kankan AI platform are almost without limit.  We have not yet even scratched the surface of what is being achieved with Kankan.  

For sure I should disclose and re-empahsize I am long Remark Holdings.  

I suspect it will be the best investment I have ever made. 

Please be clear that this note does NOT represent investment advice or guidance.  I am simply sharing my own personal perspectives which I feel are well based, qualified and in my wheelhouse. 

Sina Weibo – http://ir.remarkholdings.com/releasedetail.cfm?ReleaseID=1044877

Sina Weibo – http://ir.remarkholdings.com/releasedetail.cfm?ReleaseID=1043838

ACXIOM http://ir.remarkholdings.com/releasedetail.cfm?ReleaseID=1046136

Alibaba AI award – http://ir.remarkholdings.com/releasedetail.cfm?ReleaseID=1042145

Anti piracy seven figure deal – http://ir.remarkholdings.com/releasedetail.cfm?ReleaseID=1041973

Food safety seven figure deal – http://ir.remarkholdings.com/releasedetail.cfm?ReleaseID=1041791

Face++ money raise valuation – https://www.reuters.com/article/us-tesla-results/tesla-stock-falls-on-model-3-delays-biggest-ever-quarterly-loss-idUSKBN1D15TP

Disclaimers & Disclosures: For a full list of disclaimers and disclosures, please visit: www.tailwindsresearch.com/disclaimer/


  1. Their Q2 financials show that they have negative working capital of approx $65MM – scary. Any solution coming?
    They are also in default on their financial covenants and loans could be called anytime – have they received a waiver yet for those?

    • They have two potential solutions to the debt problem. ShareCare monetization and, possibly, sale of a piece of KanKan. Spoke with CFO and he’s not at all concerned about the debt situation.

  2. I wouldn’t expect the CFO to express any other reaction. But as of June 30, that 65MM WC shortfall is due within a year and another quarter has gone by and there haven’t been any announcements – the clock is ticking.

    • Agreed…but, I think you’d also agree that the valuation is reflective of this. Asset value of the Company far exceeds enterprise value.
      They get something done regarding the debt and it will be a large positive.

  3. what interests me and i am just trying to reason is the peer group of remark holdings. so far I would count splunk, hortonworks, myomo. what do you think?

  4. Thanks. When i read something like this about sensetime “It is seeking to expand overseas” it looks maybe for future listing.
    Do you have ideas for actually listed AI-firms?

    • GSUM is a good one to look at. I’m not aware of too many pureplays. VERI stock did very well recently when people realized they are an AI company…this is the kind of performance I think we can get from MARK, only Remark is a better company.

      • VERI has a spectacular performance behind it, that’s right. GSUM is worth a look, let’s take a closer look. * shake hands * – the attention is so far very low regarding Mark. If they are attentive then it goes off as Mark has in recent days. I hope for a little reset. Mark has everything in front of him, and we agree on that. I will continue to search and post this if I find more companies. Why do you think Mark is better than the ones mentioned so far (also because of the good cooperation?)?

        • MARK has two things going for it. Access to Alibaba data gives them a data set of 1.3 billion people. That’s unreal.

          Secondly, their platform is very well constructed which gives them the ability to launch new products in short time frames. This is a key differentiator. Cost and time matter and they are better on both.

          • This Ali Cooperation is the most important advantage i think. For example I’m impressed by the service of aliexpress compared to Amazon, they’re a better (and faster) one.
            That’s really interesting too.