Companies store a lot of unstructured information and often do not have the capacity to make much of it.
Now think you have a way of storing information and you can really ask questions, such as, “When did ABC Company sign its first contract with us?” Or “Show me videos with blue skies”
That’s what CEMI Technologies is developing with Weaviate, a vector search engine. Using machine learning models is a unique type of AI-first database of vectors output, also known as embeddings, so it has been dubbed a vector search engine, ”said Bob Van Louiej, CEO and co-founder of Semi.
He explained that vector search engines are not new – Google Search is an example of a solution built into a vector search engine. But SemiI’s goal is to pack this technology and have an open source business model for anyone to use.
Last year, Van Lujojet created a question-and-answer translation program for my colleague Alex Wilhelm on the 2021 Techcrunch articles under the technology cover last year.
“Everyone can use the technology, and we have the tools and services for the companies that need it,” added Van Louis. We do not create or distribute real models – this is what companies like Huggingface or OpenAI do or companies make models themselves. But it is one thing to have the models, it is quite another to use them in your search and consulting services in a product, and this is what Wewitt solves.
The company was founded in Since its inception in 2019 by CTO Etienne Dilocker and COO Micha Verhagen, Van Lujjet has explored more than 100 issues using semiconducting technology, including start-ups such as Kinus or Zenkaster, based on new options offered by a vector search engine. And uses the results provided by Weaviate directly to help people, for example in medicine.
Some of Van Louis’s personal favorites can easily search the human genome, including the ones that are “very messy”, including the genome vector verification and exploration, the map or graph imaging of the whole world. Semi Sequence with Weaviate as created on Meta Researches graphs.
SemiI raised $ 1.2 million from Zeta Venture Partners and ING Ventures in August 2020 and has been on Venture Capital companies’ radar ever since. Since then, the software has been downloaded about 750,000 times, an increase of about 30% per month. Van Louis did not elaborate on the company’s growth metrics, but said the number of downloads could be related to enterprise licenses and sales of managed services. In addition, the use and awareness of Weaviate Value Added all growth metrics and the company wasted its cash flow.
Although the seed was lost, the company did not seek new funding. However, while SemiI co-founders enjoyed talks with Cortical Ventures, a new fund from former databases and New Enterprise (NEA) founders, Van Louis, showed the companies how they support the business.
He added: “It was wonderful to ‘pinch my hand to my jaw.’ “Everything he has done in the past has been exactly what we want the support team to do, and although he has a very new experience, all the amazing stories are true.”
Those talks led NEA and Cortical to a new round of $ 16 million in Serie A.
SemiI aims to recruit new US and European talents and double the open source community for search and vector search as a whole. It will also focus on marketing and open source products around open source cores, and will take the first steps in the development of machine learning in conjunction with computer science.
Meanwhile, Van Louis believes he is looking at the next database technology wave that started the SQL wave that led to big winners such as Oracle and Microsoft, while the second wave was a no-SQL database wave, winners like MongoDB and Redis.
“We are now on the verge of a new database, these are AI-starters, and Weaviate is an example of this,” he added. “We have to educate the market not only about Weaviate but also about vector search databases or AI-start databases. This is exciting because machine learning brings something unique to the table. Reply or ‘understand’ what millions of photos or videos contain.