Gem Spotlight - Vectorspace AI
Today’s newsletter is going to be a breakdown of the project ‘VectorSpaceAI’, I previously got asked many times what projects I was interested in and investing in, so I thought I would start producing this style of newsletter to analyze some exciting projects that I thought looked interesting in terms of both fundamentals & technicals.
If you do enjoy this newsletter then please make sure you sign up. I write newsletters a few times a week discussing the markets and talking about my technical analysis which I hope can be useful if you're looking for more detailed reports. Thanks guys!
In this specific newsletter, I wanted to take a deep dive into ‘Vectorspace AI’, a unique platform that is being built inside of the blockchain ecosystem. Vectorspace AI is focused on context controlled NLP/NLU (Natural language processing/ understanding) & feature engineering for hidden relationship detection.
Vectorspace AI got its start in the Life Sciences and bioinformatics area while at Lawrence Berkeley National Laboratory. Today, the group algorithmically generates datasets that power machine learning systems to identify hidden relationships between approved drug compounds for repurposing applications in fighting COVID-19 along with relationships between global events and stocks that exhibit high correlations to each other (a form of automated thematic investing). These relationship networks (visualized below) are based on proprietary datasets which in turn are built using natural language processing (NLP) and natural language understanding (NLU) technology.
I’m particularly impressed by the quick adoption to the overall circumstances when COVID-19 just began spreading. Vectorspace AI was one of the first movers to provide their services for free and announce the availability of real-time COVID-19 (Coronavirus SARS-CoV-2) drug repurposing datasets in collaboration with Amazon and Microsoft in connection with the United States Office of Science and Technology Policy (OSTP).
"One of the most immediate and impactful applications of AI is in the ability to help scientists, academics, and technologists find the right information in a sea of scientific papers to move research faster. We applaud the OSTP, WHO, NIH and all organizations that are taking a proactive approach to use the most advanced technology in the fight against COVID-19," said Dr. Oren Etzioni, Chief Executive Officer of the Allen Institute for AI.
To give a real-life example on how we could effectively use these datasets, Vectorspace AI currently has opened to the public a COVID-19 dataset builder that allows users to get real-time datasets for hidden relationship detection. This accelerates the speed of innovation and novel scientific breakthroughs, both of which are crucial to drug development and repurposing in the fight against COVID-19. The potential of Vectorspace AI is limitless as it is a unique offering being built in an emerging and complex market. This is why, in this newsletter, I am excited to dig into the fundamentals and highlight why projects stands out.
Firstly, I would just like to mention that generally speaking, this is a pretty underdeveloped area of the blockchain space. As such, I have to give credit to the team for working on something that is so truly unique and complex. Something that could have a large impact on the world from a scientific standpoint. Any product/platform that has the goal of pushing forward the boundaries of technology and science is something to be admired, and I am excited to see how this platform rolls out.
“We are particularly interested in how we can get machines to trade information with one another or exchange and transact data in a way that minimizes a selected loss function. Our objective is to enable any group analyzing data to save time by testing a hypothesis or running experiments with higher throughput. This can increase the speed of innovation, novel scientific breakthroughs and discoveries”
Kasian Franks is the technical founder of Vectorspace AI and brings extensive experience as a life sciences researcher, hedge fund advisor, ML/AI data engineer, and serial entrepreneur. Equipped with decades of relevant and extensive experience across a breadth of fields, Kasian Franks has all the necessary tools to lead and inspire an ambitious project such as this.
The Vectorspace AI team is composed of industry experts in the science, data, and financial markets. Primarily made up of veteran software engineers, scientific experts, and technical founders, their experience in developing products and applications within the AI/ML field harkens all the way back to 1994.
From top to bottom, the team is profoundly experienced and well-rounded in Artificial Intelligence & Machine Learning, making them more than capable of delivering a game-changing product such as Vectorspace AI successfully to market. Although the team composition has been lean through the initial development of the platform, the company is rapidly expanding to match the production needs of the customer onboarding that is even now taking place. I personally believe the team has already delivered such an incredible product and the amount of experience within the team itself makes me confident that no matter the size, they have the correct knowledge and insight to really build this platform to be something truly special.
Vectorspace AI Datasets
Vectorspace AI offers a wide range of products/features inside of their datasets, ranging from Dataset augmentation, Real-time insights, and Built-in data provenance to a State-of-the-art data pipeline. While the Vectorspace AI NLP/NLU datasets are endlessly customizable and can be leveraged by any industry, Vectorspace AI also provides access to many high-quality data sources out-of-the-box for anyone to use.
Vectorspace AI provides data augmentation services in the form of static and real-time, context-controlled, correlation matrix datasets based on Natural Language Processing (NLP) and Natural Language Understanding (NLU). The datasets can be applied in all industries to generate new interpretations, hypotheses, and discoveries.
Using the API, Vectorspace AI customers have access to near-real-time (NRT) datasets that update as frequently as once per minute (1440 API calls per day), allowing for correlation scores and insights that can be generated in isolation, or as an augmentation to any external or internal dataset.
Built-in data provenance
For advanced users, Vectorspace AI offers optional data provenance solutions via their Data Provenance Pipeline (DPP). The DPP rigorously controls data lineage, ensuring that users always know exactly where the data originated from and how it was processed. This is a must-have for bioscience and financial institutions that rely on their datasets to make billion-dollar decisions every day.
State-of-the-art data pipeline
The Vectorspace AI data engineering pipeline takes unstructured text from any data source and applies state-of-the-art machine learning techniques based on unsupervised learning and NLP/NLU to find hidden relationships between entities (like genes or stocks) that can provide a valuable “signal” for their customers.
While I’m personally not experienced in the field of datasets and scientific studies, it is very clear from looking at the list of optional features above that Vectorspace AI has thought about the two different types of end-user; Firstly the one who is very experienced in a certain field and needs data for expert-level studies, and secondly, the less experienced user who needs to use a pre-built dataset to access information.
Based on the products and features, the platform is clearly geared more towards educational and expert users alike. Additionally, I would like to highlight the abundance of documentation and articles Vectorspace AI provides on the use of the platform, datasets, and education of ML/AI – All of which are great for users interested in learning more. At a high level, datasets can be controlled and held up by large corporations or entities who leverage the data to make strategic business decisions, find alpha, or reduce cycle time in research studies. Vectorspace AI is allowing anyone and everyone to be able to access datasets they need, on a safe, secure, and fast platform.
I wanted to share some use-cases for Vectorspace AI and demonstrate some of the content that they are providing on their website for users to gain an understanding of when and how the platform can be used. First off, let's take a look at the potential use cases for NLP based correlation matrices.
Create unique networks of clusters based on concepts and hidden relationships and compare their gains to those of S&P 500
Determine if price correlations have a similar concept or keyword correlations
Examine symbiotic, parasitic, and sympathetic relationships between equities
Create within seconds baskets of stocks based on concepts, keywords, and/or a link to a news article (information arbitrage)
Use stock symbols as custom concept column labels and model cross-correlations between equities
Create features using trending terms anywhere on the internet
These are just a few examples of how Vectorspace AI’s datasets could be used. As you can see there are a number of different ways in which NLP-based correlation matrices could be used in regards to the S&P 500 and stocks, which is something most people who are reading this newsletter are familiar with. So, you could also imagine how something like this could be tied into cryptocurrency markets and other blockchain elements.
An approved drug is discovered to have a new application for a different disease (drug repositioning/repurposing) - extract a cluster of other drug compounds that may have known or hidden relationships.
A drug compound passes phase 3 clinical trials for Company XYZ - cluster other public company pharmaceutical companies that may have known or hidden relationships to company XYZ.
Company XYZ has a drug compound that's found to cause significant health problems - find a cluster of other public company pharmaceutical companies that may share the same risk as company XYZ due to similar drugs being developed in their public company pharmaceutical pipelines.
To further illustrate the above example, according to an article ‘Dexamethasone Announcement Could Have Made Hedge Funds A Fortune’ published in AlphaWeek, a hedge fund industry trade publication, by Winterton (2020), “San Francisco-based machine learning technology firm Vectorspace AI has built an algorithm which constructed an equally-weighted basket of five securities containing Novavax, Inc. (NVAX), Perrigo Company plc (PRGO), Bristol-Myers Squibb Company (BMY), Karyopharm Therapeutics Inc. (KPTI) and Momenta Pharmaceuticals, Inc. (MNTA) which in two days returned +21.23% on the back of the Dexamethasone announcement; the only one of the five stocks to lose money was MNTA, which fell -1.69% in the two days following the news. NVAX returned +15.81% for the best performer award.” (Par. 2).
Note: As of August 5th, the returns by Novavax Inc. (NVAX) is at 3733.333%.
As you can see from the example above, smart baskets around pharmaceuticals illustrate one of the many ways in which Vectorspace AI can have an impact on the world. The possible use cases for a product like Vectorspace AI are unlimited and I am really excited to see what people will build and develop while using the data that can be created and analyzed on the platform.
The Vectorspace AI platform and product offerings are unique, giving it a first movers’ advantage in a field that is underdeveloped and ready to skyrocket in the blockchain space. It is exciting to see new and interesting ways in which blockchain applications are being built, and I personally love what Vectorspace AI is building.
Use cases for the product are vast and encompass everything from uncovering hidden relationships between diseases and pharmaceuticals to driving academic research and finding alpha in the financial markets. Having a large variety of use-cases means that Vectorspace AI has a great deal of potential to build a wide user base that covers many different elements and sectors. This is something you want to see in new projects as it drives growth and makes it easier to rapidly scale user adoption.
Overall, this seasoned team possesses both the business acumen and technical expertise to develop the Vectorspace AI platform into something extraordinary. The combination of a versatile use-case, embedded token model, and profit-first strategy presents a singular incentive for investors to buy in early and ride this new wave. I am very excited to see over the coming weeks and months how they rollout, and I am very excited to see them onboard more users and gain market share in this new sector.
If you enjoyed this newsletter please make sure to sign up with your email address, it's free and I write newsletter content a few times a week! I really enjoy writing these articles! If you would like to find out more about Vectorspace AI, take a Deep Dive on who they are. You can head over tohttps://vectorspace.ai, join their Community Channel on Telegram or find them on twitter at @Vectorspace_AI. Thanks again.
Winterton, Greg. “Dexamethasone Announcement Could Have Made Hedge Funds A Fortune.” AlphaWeek, AlphaWeek, 20 July 2020, www.alpha-week.com/dexamethasone-announcement-could-have-made-hedge-funds-fortune.