The advent of generative artificial intelligence has made data management a pressing concern, with data software companies looking to better support the average business user. Founded in 2021, Illumex Technologies Inc. is set to meet businesses’ data needs through its Generative Semantic Fabric, an abstraction layer that automates data mapping and analytics.
“The biggest flaw so far in data management in big organizations is actually that technical people are managing data. Because we have the data to enable business in the first place, but there is [a] huge disconnect between business and data and what we need [is] more business people helping to maybe govern and certify and be involved more in the semantic,” said Inna Tokarev Sela (pictured), founder and chief executive officer of Illumex. “But it doesn’t have to be on the technical level. It should be enabled to be [an] application interface.”
Tokarev Sela spoke with theCUBE Research’s John Furrier at the Supercloud 7: Get Ready for the Next Data Platform event, during an exclusive broadcast on theCUBE, SiliconANGLE Media’s livestreaming studio. They discussed how Illumex is transforming enterprise data management and why AI is changing the data analytics space
What Generative Semantic Fabric brings to data management
While neural networks imitate the human brain, knowledge graphs—Tokarev Sela’s specialty—provide context for a user’s data. With its Generative Semantic Fabric, Illumex helps business users better understand and harness data through a unifying, contextual layer of abstraction.
“GSF is the layer which incorporate[s] content and context. So it’s a knowledge graph of semantic embeddings which sits between data and applications,” Tokarev Sela said. “It understands two sides of the table and it allows business users to interact in free language, not in data language with any of the data connected to it, and it’s translated to the right assets with [a] high degree of governance and explainability.”
Data management has experienced a resurgence because of AI but a lot of data remains siloed, and there continues to be a lack of data science experts. Illumex’s semantic layer is aimed at mapping AI-ready data and creating wider access to data with agentic AI, according to Tokarev Sela.
“Big problems which we solve [are] … AI-ready data, being able to map all the data and distill it to business language, basically map it to semantic language,” she said. “And the second part is connecting it to the user input … this translator which comprehends a human impact and free language and maps it to this knowledge graph of semantic embeddings.”
Companies were not necessarily motivated to break data silos until gen AI presented a “killer” use case, according to Tokarev Sela, which has now created a need for a semantic layer that connects entire organizations.
“We have to have [a] single source of truth [for gen AI] and silos of course prevent us from doing so. And this is why generative AI is the ultimate use case to break silos and finally have unified semantics across organization and even unified semantics across the whole data state,” she said. “So again, if you use external data, APIs and so on and so forth, you’ll be able to speak the same language to all of those data sources.”
Tailoring Generative Semantic Fabric to the business
Even as companies work toward solving the data silo problem, their data stores will still be geographically dispersed. However, Tokarev Sela does not see connecting to different data sources as a problem because Illumex’s semantic fabric will be trained on a business’s unique terminology.
“Generative Semantic Fabric [is], in essence, exactly this glue which connects to the metadata of all organizational structured data sources and business applications … we automatically build it in a coherent terminology, business glossary, metrics store, lineage and all of that,” she said. “We also wrap it up with [an] augmented governance process … we are picking up all [the] different definitions from those silos, those different functional silos and maybe geographical silos.”
A company’s data should be available to the whole business, Tokarev Sela believes, so working toward that contextualizing layer will ultimately drive more business value for “data-intensive” industries such as pharmaceuticals and financial services.
“Graphs capture a concept and they also capture the relationship between concept. But what makes it special for every company is they have own business context concepts which are not captured in general purpose algorithm,” she said. “So this is where it’s really [a] differentiator to implement those in graphs … it’s the fastest way to receive information.”
Stay tuned for the complete video interview, part of SiliconANGLE’s and theCUBE Research’s coverage of the Supercloud 7: Get Ready for the Next Data Platform event.