--- title: "What Is Illumex Business Context AI? Inside the NVIDIA Deal" url: "https://atlan.com/know/ai-agent/illumex-business-context-ai/" description: "Illumex built a Generative Semantic Fabric for business context AI before NVIDIA bought it in 2026. See what it did and how it compares to a full context layer." author: "Emily Winks" author_role: "Data Governance Expert" published: "2026-08-11" updated: "2026-08-11T00:00:00.000Z" --- --- Illumex was an Israeli startup that built a "Generative Semantic Fabric," software that auto-generated business-meaning definitions from structured-data metadata. NVIDIA acquired the company in February 2026, [reportedly for $60 million to $75 million](https://www.jpost.com/business-and-innovation/tech-and-start-ups/article-887707), and Illumex no longer operates as an independent company. Atlan's context layer treats that same job, generating business meaning, as one of three substrates it covers, not the whole platform. --- Illumex was founded in 2021 in Tel Aviv by CEO Inna Tokarev Sela. It [raised a $13 million seed round in June 2024](https://www.crunchbase.com/organization/illumex) and counted Teva Pharmaceutical among its named users. It already had a working NVIDIA relationship, a NeMo integration announced in March 2025, before the acquisition, so the deal extended a partnership rather than starting cold. - **What it did well**: metadata-only, privacy-preserving semantic mapping that never touched raw row-level data - **What it didn't solve**: a product tied closely to NVIDIA's own AI stack, not built to be model-agnostic - **What the rest of this page answers**: whether "business context AI" is a standalone category, or one substrate inside a larger context layer | Field | Detail | |---|---| | What it was | A "Generative Semantic Fabric" platform that auto-generated business-meaning definitions from structured-data metadata | | Founded | 2021, Tel Aviv, by CEO Inna Tokarev Sela | | Funding raised | $13M seed round, June 2024, led by Cardumen Capital with Amdocs Ventures and Samsung Ventures (reported lifetime totals vary by source) | | Acquired by | NVIDIA, reported February 23-25, 2026, for $60M-$75M (sources vary; both figures are reported, no single authoritative number) | | Named customers/partners | Teva Pharmaceutical, Carson Group; partnerships with AWS, Google Cloud, Microsoft, and NVIDIA | | NVIDIA-stack dependency | Generative Semantic Fabric was already tied to NVIDIA's own AI stack (a NeMo integration announced March 2025) before the acquisition made that tie permanent | --- ## What is Illumex, and why did it call itself "business context AI"? The substrate Atlan's Semantics & Ontology layer covers is the same one Illumex built its entire company around. Illumex was an Israeli software company that built a product it called a "Generative Semantic Fabric," designed to auto-generate business-meaning definitions from structured-data metadata rather than from the underlying data itself. Founded in 2021 in Tel Aviv by CEO Inna Tokarev Sela, the company raised a $13 million seed round in June 2024, led by Cardumen Capital with participation from Amdocs Ventures and Samsung Ventures, according to [SiliconAngle](https://siliconangle.com/2024/06/27/illumex-raises-13m-add-meaning-context-structured-data-used-generative-ai/) (2024). That $13 million round is the most consistently documented figure across coverage of the company; reported totals for Illumex's lifetime funding vary by source. "Business context AI" was Illumex's own term for what its product did: read the schema and column-level metadata sitting on top of structured data, then infer what those columns and tables actually meant to the business, without moving or exposing a single row of the source data. Tokarev Sela described the goal directly: "Our Generative Semantic Fabric aligns organizational data with business meaning and domain-specific context, allowing organizations to finally trust the results of their AI initiatives," according to [SiliconAngle](https://siliconangle.com/2024/06/27/illumex-raises-13m-add-meaning-context-structured-data-used-generative-ai/) (2024). That is the same underlying job Atlan's [Semantics & Ontology](https://atlan.com/know/what-is-context-layer/) foundation layer does: generate shared business definitions instead of leaving each tool to define them alone. A definition that is generated once and lives in a single tool is still a fragment, not a fix, for the reconciliation problem the rest of this page works through. --- ## How did Illumex's Generative Semantic Fabric actually work? The mechanism behind Illumex's core claim was automated, metadata-only semantic mapping. The Generative Semantic Fabric analyzed schema, column names, and usage metadata, never the raw row-level data, and used that to infer what a table or column meant and how it related to others. Gonzalo Martínez de Azagra, founder and general partner at Cardumen Capital, Illumex's lead seed investor, pointed to data integrity as the reason he backed the company: "Illumex perfectly bridges this gap, enabling enterprises to maximize the GenAI opportunity while ensuring the integrity, lineage, and trustworthiness of the data fueling these applications," according to [PR Newswire](https://www.prnewswire.com/news-releases/illumex-announces-13m-funding-to-enable-trustworthy-and-governed-enterprise-genai-with-structured-data-302184350.html) (2024). Illumex has said the approach worked at scale. On an unnamed retailer's ERP migration spanning 115,000 tables and 1.5 million data columns, it reported cutting a projected two-year manual review to one week of automated workflows, according to [Illumex's own account, published via PR Newswire](https://www.prnewswire.com/il/news-releases/illumex-accelerates-digital-transformation-for-erp-systems-with-nvidia-ai-software-302404643.html) (2025). That's Illumex's claim about its own deployment, not an independently audited result. ### What Illumex's own ERP-migration case shows and doesn't show The case is a real, credible demonstration that automated semantic mapping can compress a painful manual process. What it doesn't show is how those definitions were governed afterward, whether they stayed current as source systems changed, or who owned correcting one once a schema shifted. A single vendor-reported case study proves the generation step works; it says nothing about the maintenance step after it, and that gap is where a fragmented, per-tool approach to meaning tends to break down. Get the CIO's Guide to Context Graphs A framework for deciding what your AI agents actually need before you buy a semantic tool, a platform upgrade, or a full context layer. Get the Guide --- ## Why did NVIDIA acquire Illumex in 2026? This section lays out the deal itself: what happened, when, and why. Multiple independent outlets reported that NVIDIA acquired Illumex between February 23 and 25, 2026 (the range likely reflects a timezone difference, not a disputed fact), for a price reported between $60 million and $75 million. [The Jerusalem Post](https://www.jpost.com/business-and-innovation/tech-and-start-ups/article-887707) reported the higher figure, while other outlets reported the lower one; no single number is authoritative, so both are worth reporting. Illumex's team joined NVIDIA's Israeli R&D organization as part of the deal. The acquisition wasn't a cold start. The two companies had already announced a partnership integrating the Generative Semantic Fabric with NVIDIA NeMo for ERP-focused generative AI in March 2025, according to [R&D World](https://www.rdworldonline.com/how-illumex-and-nvidia-are-partnering-to-address-the-trust-problem-with-genai-and-agents/) (2025), which is the working relationship the acquisition formalized rather than replaced. ### Timeline: from a $13M seed to an NVIDIA acquisition | Date | Milestone | Detail | |---|---|---| | 2021 | Founded | Tel Aviv, by CEO Inna Tokarev Sela | | June 2024 | $13M seed round | Led by Cardumen Capital, with Amdocs Ventures and Samsung Ventures | | March 2025 | NVIDIA partnership announced | Generative Semantic Fabric integrated with NVIDIA NeMo for ERP-focused generative AI | | Feb 23-25, 2026 | NVIDIA acquisition reported | $60M-$75M (sources vary); team joins NVIDIA's Israeli R&D organization | ![Illumex timeline from startup to NVIDIA: founded in Tel Aviv in 2021, a $13M seed round in June 2024, an NVIDIA NeMo partnership in March 2025, and NVIDIA's acquisition of Illumex in February 2026](/img/illumex-business-context-ai-1-nvidia-acquisition-timeline.webp){width=1672 height=941} A five-year journey from Tel Aviv startup to NVIDIA acquisition. Source: Atlan. A real acquirer paying real money for a narrow, well-executed semantic tool is evidence the underlying problem, business meaning that's generated once and actually trusted, is worth solving. It isn't proof that solving meaning alone is the whole job, though NVIDIA folding the team into a broader stack rather than reselling the product standalone is one signal in that direction. The next section covers exactly what that folding-in looked like. --- ## What happens to Illumex's product now that NVIDIA owns it? Of everything written about Illumex, what actually happened to it after the deal closed has the least existing coverage anywhere online. Illumex no longer operates as an independent company and doesn't sell the Generative Semantic Fabric as a standalone product. Its own domain, illumex.ai, no longer resolves as of this writing (confirmed by a direct check, not a claim from any third party), and its [G2 reviews page](https://www.g2.com/products/illumex/reviews) shows a thin, dated review base, consistent with a wound-down brand. The team and technology were absorbed into NVIDIA's Israeli R&D organization, folding an acquired capability into a broader agentic AI stack rather than running it as a separate business. Illumex's technology was already tied closely to NVIDIA's own stack before the deal, a narrower commitment than a [model-agnostic context layer](https://atlan.com/know/ai-agent/context-layer/model-agnostic-context-layer/) built to work the same way regardless of which model or vendor sits on top. For a reader evaluating tools today, the question shifts. It's no longer "should we buy Illumex," since that option doesn't exist. It's "what does a point solution for this one substrate look like, and is it enough on its own." NVIDIA's choice to integrate rather than resell doesn't settle that outright, but it's a data point worth weighing: a well-executed acquirer treated generated meaning as one ingredient in a larger stack, not a sellable end state by itself. Try the Context Gap Calculator See how much of your data estate a single semantic tool actually reaches, and where the gaps are. Calculate Your Gap --- ## Is a semantic layer the same as a context layer? This section answers the structural question the rest of the page has been building toward. A semantic layer, Illumex's Generative Semantic Fabric among them, defines what things mean: shared business definitions and the relationships between them. Atlan's [context layer](https://atlan.com/know/what-is-the-enterprise-context-layer/) treats that job, semantics and ontology, as one of three integrated substrates, alongside trusted, AI-ready data and the skills and procedures for how work gets done. Illumex itself never marketed "business context AI" as a full context layer; it was built to do the semantics job well, not the other two. A tool that only does that first substrate answers "what does this mean" well, but has nothing to say about whether the data can be trusted or how an agent should act once it has the definition, which is a scope difference rather than a shortcoming in what Illumex set out to do. A tool built tightly around one AI vendor's stack is also harder for a broader agent ecosystem to consume, regardless of how good its definitions are. Illumex's technology was closely tied to NVIDIA's own AI stack, an integration with NVIDIA NeMo announced in March 2025, months before the acquisition made that relationship permanent. Atlan ships an [MCP](https://atlan.com/know/mcp-delivers-business-context/) server today, built to work the same way regardless of which model or vendor sits on top; it reflects an architectural choice to make context reachable by any agent instead of tied to one vendor's stack, the same choice examined in [context layer vs. knowledge graph](https://atlan.com/know/ai-agent/context-layer/context-layer-vs-knowledge-graph/). --- ### The three substrates of a context layer | What it covers | Illumex's Generative Semantic Fabric | Atlan's context layer | |---|---|---| | Semantics & ontology | Yes, its core, well-executed function | Yes, one of three substrates | | Trusted, AI-ready data | Not its focus | Yes, a separate, integrated substrate | | Skills & procedures | Not its focus | Yes, a separate, integrated substrate | | Model-agnostic delivery | No, tied to NVIDIA's own AI stack | Yes, ships an MCP server that works with any model or vendor | | Governed as the business changes | Not documented | Yes, active governance, not a one-time pass | A generative semantic layer and a context layer solve related but different-sized problems. Confusing the two is how the same metric ends up defined five different ways across five well-intentioned tools that don't know about each other. --- ## How does Atlan approach business context differently? Atlan's own approach to the same job Illumex did goes further, and here's where. Auto-generating business definitions from metadata is a real, valuable capability, and Illumex earned real credit for it: a working product, credible enterprise users, and [NVIDIA paying $60 million to $75 million for it](https://www.jpost.com/business-and-innovation/tech-and-start-ups/article-887707), proof the market agrees the underlying problem is worth solving. But a definition generated once and never governed, kept current, or served safely to every agent that needs it is only half the job. Atlan's Semantics & Ontology substrate does the same core generation work, integrated with trusted, AI-ready data and skills and procedures, delivered through an open MCP server built to work the same way regardless of which model or vendor sits on top. One current example: Atlan's [Snowflake semantic views](https://atlan.com/know/snowflake/snowflake-semantic-views/) sync keeps a definition consistent between where it's authored and where analysts and agents query it, instead of letting the two drift. The same substrate is examined in [semantic layer for AI agents](https://atlan.com/know/ai-agent/semantic-layer-for-ai-agents/), and the broader pattern of a platform folding in a point solution is covered in [full-stack AI platform vs. best-of-breed context layer](https://atlan.com/know/ai-agent/context-layer/full-stack-ai-platform-vs-best-of-breed-context-layer/). ### What a real customer said, unprompted > "I expected boilerplate; instead, it inferred business context I never explicitly provided, correctly describing how an asset fit into our customer journey just from column names and lineage. That's when I realized this was knowledge synthesis, not just documentation." > > Swatilekha Saha, Data Architect, DAT That's the same outcome Illumex marketed as "business context AI," described independently by an Atlan customer who reached for the same language unprompted, which carries more weight than a marketing claim for the obvious reason that nobody asked her to say it. [Business context for AI](https://atlan.com/know/business-context-for-ai/) covers that same territory from Atlan's side directly. See Atlan in Action: Live Context Layer Demos See how Atlan generates and governs business context as one part of a full context layer, not a standalone tool. Watch a Live Demo --- ## Is business context AI a category, or a feature of something bigger? Illumex's acquisition answers half of the question a reader searching this term actually has, and it's worth being honest about which half. [NVIDIA paying $60 million to $75 million](https://www.jpost.com/business-and-innovation/tech-and-start-ups/article-887707) for a semantics-only point solution proves the underlying problem, business meaning generated once and then trusted, is real and worth acquiring. It doesn't prove that solving meaning alone is a complete answer, and NVIDIA folding Illumex into a broader stack rather than reselling it standalone is a signal in that direction, not a settled proof. Business context AI, as Illumex defined it, is a real, valuable capability the company never claimed was the entire context problem. It just isn't the whole job on its own, which is a scope question, not a knock on what Illumex built. Book a Demo --- ## FAQs about Illumex and business context AI ### 1. Is Illumex still an independent company? No. NVIDIA acquired Illumex in February 2026, reportedly for $60 million to $75 million, folding its team into NVIDIA's Israeli R&D organization. Illumex no longer sells its Generative Semantic Fabric as a standalone product, and its own domain no longer resolves. ### 2. What happened to Illumex after the NVIDIA acquisition? Illumex's team and technology moved into NVIDIA's Israeli R&D organization rather than continuing as a separate business. NVIDIA hasn't repositioned the Generative Semantic Fabric as a standalone product; it's a capability inside NVIDIA's broader agentic AI stack instead. ### 3. What are the applications of Illumex? Illumex's Generative Semantic Fabric auto-generated business definitions from structured-data metadata, most publicly documented in enterprise resource planning migrations, where it inferred what tables and columns meant without moving the underlying data. Named enterprise users included Teva Pharmaceutical and Carson Group. ### 4. How much funding did Illumex raise before being acquired? Illumex's best-documented funding round was a $13 million seed round in June 2024, led by Cardumen Capital with participation from Amdocs Ventures and Samsung Ventures. Reported figures for the company's total lifetime funding vary by source. ### 5. What is a "Generative Semantic Fabric"? It's Illumex's term for software that automatically infers business-meaning definitions and relationships from structured-data metadata, such as schema and column names, without processing the underlying row-level data itself. ### 6. Is a semantic layer the same as a data catalog or context layer? No. A data catalog inventories what data exists and where. A semantic layer defines what that data means. A context layer covers both of those plus a third piece, the skills and procedures for how work gets done, unified into one governed system. ### 7. Was Illumex tied to the NVIDIA stack, or model-agnostic? Illumex's Generative Semantic Fabric was closely tied to NVIDIA's own AI stack. The two companies announced an integration with NVIDIA NeMo in March 2025, extending Illumex's product into NVIDIA's ecosystem well before the acquisition made that relationship permanent. That's a narrower commitment than a model-agnostic context layer, built to work the same way regardless of which model or vendor sits on top. --- ## Sources 1. [Illumex raises $13M to add more meaning and context to structured data used by generative AI, SiliconAngle](https://siliconangle.com/2024/06/27/illumex-raises-13m-add-meaning-context-structured-data-used-generative-ai/) 2. [Illumex announces $13M funding to enable trustworthy and governed enterprise GenAI with structured data, PR Newswire](https://www.prnewswire.com/news-releases/illumex-announces-13m-funding-to-enable-trustworthy-and-governed-enterprise-genai-with-structured-data-302184350.html) 3. [Nvidia acquires Israeli data co Illumex, The Jerusalem Post](https://www.jpost.com/business-and-innovation/tech-and-start-ups/article-887707) 4. [How Illumex and NVIDIA are partnering to address the trust problem with GenAI and agents, R&D World](https://www.rdworldonline.com/how-illumex-and-nvidia-are-partnering-to-address-the-trust-problem-with-genai-and-agents/) 5. [Illumex accelerates digital transformation for ERP systems with NVIDIA AI software, PR Newswire](https://www.prnewswire.com/il/news-releases/illumex-accelerates-digital-transformation-for-erp-systems-with-nvidia-ai-software-302404643.html) 6. [Illumex company profile, Crunchbase](https://www.crunchbase.com/organization/illumex) 7. [Illumex Reviews 2026: Details, Pricing & Features, G2](https://www.g2.com/products/illumex/reviews)