Companies sit on a lot of unstructured data and often do not have the capabilities to get much out of it.
Now imagine having a way to store data and actually ask it questions, for example, “When did ABC Company sign its first contract with us?” or “Show me videos with blue skies.”
That’s what SeMI Technologies is building with Weaviate, a vector search engine. It is a unique type of AI-first database that uses machine learning models that execute vectors, known as embeddings, hence the name vector search engine, said Bob van Luijt, CEO and co-founder of SeMI.
He explained that vector search engines are not new: Google Search is an example of a solution built on top of a vector search engine. However, the goal of SeMI is to make this technology a standard and has an open source business model so that anyone can use it.
Van Luijt gave my colleague, Alex Wilhelm, a peek under the hood of technology last year by creating a semantic search engine that answers questions about 2021 Techcrunch articles.
“Everyone can use the technology and we have tools and services for those companies that need it,” added Van Luijt. “We don’t make or distribute the actual models – this is something companies like Huggingface or OpenAI do, or companies make models themselves. But having the models is one thing, using them to power your search and recommendation systems into production is another, and this is exactly what Weaviate solves.”
Since founding the company in 2019 with CTO Etienne Dilocker and COO Micha Verhagen, Van Luijt has seen SeMI’s technology inspire more than 100 use cases, including startups, such as Keenious or Zencastr, and create new businesses based on the new technologies. possibilities that a vector search engine offers them, and applications where the results of Weaviate help people directly, for example in the medical world.
Van Luijt’s personal favorites, he said, were more ‘esoteric’, including vectorizing and searching the human genome, mapping the entire world in vectors, or so-called embedding of graphs, which are easily searchable. with Weaviate, such as a demo SeMI made on the embedding of charts from Meta Researches.
SeMI raised a $1.2 million seed from Zetta Venture Partners and ING Ventures in August 2020 and has been on the radar of venture capital firms ever since. Since then, the software has been downloaded nearly 750,000 times, a growth of about 30% per month. Van Luijt didn’t provide details on the company’s growth rates, but said the number of downloads may correlate with sales of business licenses and managed services. In addition, Weaviate’s spike in value-add usage and understanding has pushed all growth metrics up and the company exhausted its seed capital.
Although the seed funding had disappeared, the company was not actively seeking new funding. However, when the co-founders of SeMI held talks with Cortical Ventures, a new fund of ex-founders of Datarobot and New Enterprise Associates (NEA), Van Luijt said the companies showed them how they would support the company.
“It was really ‘squeeze my arm breathtaking’ amazing,” he added. “Everything they’ve done in the past, the teams that support us, was exactly what we were looking for, and I can say, although very fresh experience, all the great stories are true.”
Those talks led NEA and Cortical to jointly lead a new $16 million Series A dollar funding round.
SeMI plans to use the new funding to hire US and European talent and double the open source community for both Weaviate and vector search in general. It will also increase its focus on go-to-market and products around the open-source core, and take first steps in research where machine learning overlaps with computer science.
Meanwhile, Van Luijt believes we are looking at the next wave of database technology that started with the SQL wave that ushered in big winners like Oracle and Microsoft, followed by a second wave was the non-SQL database wave, with winners like MongoDB and Redis.
“We are now on the cusp of a new generation of databases, the ones that are AI-first, and Weaviate is an example of that,” he added. “We need to inform the market not only about Weaviate, but also about vector search databases, or for that matter AI-first databases. This is extremely exciting to do because machine learning brings something unique to the table. For example, having your database answer questions in natural language about millions – or even billions – of documents, or ‘understand’ what millions of photos or videos contain.”