Need to fast-track your pentesting? Our experts make it easy.
Need to fast-track your pentesting? Our experts make it easy.

Q&A with Our Partner Valuer, an AI Company

Valuer is a member of Cobalt’s partner network, which allows companies to offer their customers fast and flexible pentesting at a discounted rate and accelerate their business by including Cobalt’s PtaaS offering as part of their pentesting service offerings. In this interview, Christian Lawaetz, Chief Technology Officer at Valuer, talks about how the company helps customers navigate the world’s innovation ecosystems – and the role in which Cobalt plays in delivering that benefit. 

* Answers provided by Christian Lawaetz, Chief Technology Officer at Valuer

Let’s start with giving readers more info on your company. What made Valuer combine its AI-driven platform and research to help companies search?

Christian: Our philosophy began with an interest in providing customers with better insights into the startup world. In the beginning, our services were a manual process. Then gradually, while identifying a more standardized and successful delivery method, we identified the power and necessity of building our own strong data foundation.

We identified early on that startups were too busy hustling in their own markets to find value in sharing their information with us, so we started to investigate and build a global database of potentially relevant startup results. 

Then we faced the next challenge, how do you search in "unstructured" data where the majority of the meaning is text, which is where we identified AI and, more specifically, Natural Language Processing to be the solution to our ever-growing startup database. 

With this, one would have thought that our research initiatives would have become less and less relevant. However, we saw a point in our delivery funnel where data still required more quality than AI can consistently provide. Thus we landed on the understanding that the sum fusion of AI and human intelligence provides our customers with an altogether stronger solution than either one on its own.

How does your AI platform curate unique search results for each company searching?

Christian: A significant, and arguably the more interesting, part of our data is textual. Natural language Processing (NLP) algorithms are used to process that text into mathematical structures that, unlike text, can be utilized by a computer. Such mathematical systems also allow us to define concepts such as the distance between two bodies of text, such as descriptions of companies. 

In simpler terms, our system differs from a simple look-up method, where the result must be indexed with a specific value, where differing and potentially relevant results will not show up without the input of an exact value. In contrast, one might say that our search works non-binary—where it provides results based on their degree of similarity to the input parameters.

This functionality enables our AI to take a search query and process the individual language through NLP to understand the meaning and context of the sentence—identify potential results in our database that fits with that underlying meaning, at least somewhat. This enables us to see ourselves as a startup exploration or creative search tool. One which gives access to startup results you are unaware of—the hidden innovation opportunities.

What are your plans to evolve Valuer’s AI platform?

Christian: The majority of our improvements for our AI platform are in how we collect, process, and use startup data that populate our database. In the end, any algorithms require quality data to provide useful results. Therefore, we spend a large portion of our energy on improving and expanding our data sets across several parameters—such as industries, sectors, geographical regions, company stage, etc.

We have many different ways of improving and scaling data quality, most of which start with our data mining and sourcing activities, which are supported by a custom build pipeline that can process large datasets. 

Finally, we source and integrate any newer cutting-edge algorithms to serve the needs we have of mapping and searching our data. Our infrastructure is built to be flexible with these facts in mind, so we can easily integrate and improve our AI system when new algorithmic technologies arise.

How do you foresee AI changing in 2022 and beyond?

Christian: Several interesting movements are arising in AI currently, most of which center around text-to-image algorithms such as DALL-E and similar things. This move represents some exciting changes. There is an ease of adoption for everyday users, where AI technologies were previously kept under lock-and-key by Big Tech. Now it is much more common for smaller businesses or even private individuals to apply these new algorithms to their own unique use cases.

On the flip side of this, it raises the subject of privacy, the potential issues that AI can bring regarding personal data, and the weakness of the technology towards ethnic bias. 

I personally believe that a big part of AI in the future will have to be built with privacy elements at its core to ensure fair use of personal data. You can see that AI Ethics has become more of a focus as the broader public becomes more knowledgeable about AI. I also feel that AI won't truly become a ubiquitous part of society until some of the issues around privacy and bias are addressed and fixed.

How do you see AI reducing the workload in the cybersecurity industry?

Christian: AI has a lot of potential for specialized and standardized tasks. From my perspective, cybersecurity is a sector where there is a lot of potential for AI growth. There has been a slower adoption of AI for these fields like information security, where the cost of making mistakes is much higher than in other more traditional fields such as retail. But the changes are already happening, such as simulation and stress testing of systems or other similar tasks that need repetition while having a systematic and high skill level approach.

The future is notoriously difficult to predict. However, judging by the more general-purpose algorithms and systems we are seeing now in AI, it's not unlikely that more and more cybersecurity tasks will be handled by a complete system—one which can maintain most tasks by itself and raise a flag when it's necessary for human specialists to get involved. 

Lastly, why did you see value in a partnership with Cobalt? 

Pentesting, data management, and security go hand in hand. It was a no-brainer that partnering with Cobalt made sense for our company. They provide fast and detailed pentests to our customers and we can, in return, help their customers with AI data management. It’s top of mind for anyone in their security and compliance journey to be a part of Valuer and Cobalt.

New call-to-action

Back to Blog
About Ashley Gray
Ashley Gray is the Demand Generation Marketing Manager supporting partners at Cobalt. She is a marketing professional with experience in the Salesforce ecosystem, cybersecurity, financial services, and in various B2B realms. She is passionate about branding that supports internal teams, prospects, partners, and clients. More By Ashley Gray