Inna Tokarev Sela, CEO and founder of Illumex, is changing how the company prepares structured data for generating AI. Illumex enables the tissue to transform the secret and hidden data into meaningful, rich in commercial languages with built -in governance, and analyze the agent deployment of Genai.
The platform automatically analyzes metadata to find and mark the structured data without moving or changing it, add semantic meaning and aligned definition to ensure definition and transparency. By creating business terms, it is recommended to indicate indicators and determine potential conflicts, Illumex ensures the highest standard data governance.
Using Illumex, analyzing the agent can accurately explain user inquiries, thereby providing accurate, context perception, and no hallucination response. Under the leadership of Inna, Illumex has set a new benchmark for AI to prepare to help companies release the entire potential of its data.
What inspired you to find the reason for Illumex, how did your experience in Siseense and SAP affect your vision of the company?
During my research, I have a vision for Illumex. I think I can access related data directly without a lot of human consultation with a similar database access information.
I taught me the basis for establishing corporate software and expansion operations on SAP. Through the SAP HANA cloud platform and business plans such as the entrepreneurial partnership framework, cross -product development has enabled me to understand the needs of corporate customers. It reveals how the company handles the significant gap between the data practice and the actual needs of the end user.
On Sisense, the establishment of AI practice from scratch has proved that AI can bring huge value to customers. Seeing this influence, coupled with the rise of SaaS and Genai technology, I am convinced that my timing was to launch the correct timing of Illumex in 2021.
Illumex focuses on the semantic structure. Can you explain the core concept? What prompts you to deal with this specific challenge in AI and data analysis?
Illumex’s pioneering semantic structure-a platform that can automatically create a readable organizational environment and reasoning for humans and machines. The platform unifies the experience of sharing environment and non -technical users based on LLM -based AI and business applications.
This single structure brings two main benefits: it simplifies data management through up to 80 % of the data engineering task automation and enable non -technical users to access analysis with built -in governance, interpretability and accuracy. Both benefits have solved the billions of dollars in corporate decisions.
It is regarded as a digital playground, machines, humans, and applications spontaneously interacting without pre -programming. This is in line with our vision of non -applied future. In this case, you only need to express your tasks instead of taking into account a variety of tools, such as tables, analysis, financial systems and customers, but seamlessly complete. Givening semantic structure is the basis of this future.
What major challenges do you face in Illumex, and how do you overcome them?
In 2021, although the AI semantic model has been generated since 2017, and the graphic neural network has already existed for longer, it is a difficult task to explain to VC that we need to automatically contends and reasoning. It was a difficult task even if it was defined at that time.
What I want to say is that the biggest challenge is to really stimulate people’s excitement for this future technology and future market. I am fortunate to see my forward -looking investors.
How does ILLUMEX make organizations be AI-Ready, and why is this transition is essential in today’s business structure?
The business community is divided into two camps: the company that knows and uses AI is a transformation force similar to the Internet, and those companies that miss or delay this opportunity.
Illumex encounters organizations anywhere in the AI journey. In order to generate AI implementation, enhance and manage the logic of organizational logic and context, we realize the deployment of agency analysis and arrangement.
We have enhanced the landscape of any company for the full -stack Genai implementation platform for structured data to effectively use these advanced technologies.
Illumex emphasizes the AI response generated by “no hallucinations”. How to ensure certain and reliable output?
ILLUMEX is based on the previous business-based body theory-knowledge map capture processes, workflows and across pharmaceuticals, retail and manufacturing, as well as financial, human resources and supply chain business functions.
When boarding customers, we will automatically review these ontology on its metadata. Within a few days, the company can search for its data, verify the results and determine problems such as duplicate or conflict.
Agentic Analytics Chatbot provides a complete transparency-showing how to explain and map the problem to the client body, and then map it to the data. The combination of this transparency with automatic data verification can ensure the answer of certainty and no hallucinations. In addition, the governance team can pre -verify potential response because the context embeds all possible problems and its arrangement in advance.
How does IlluMeX distinguish between traditional methods (such as retrieval generation (RAG))?
Although RAG tries to define the ready -made AI model by feeding organization data and logic, it is facing several limitations. This is a black box-you can’t determine whether it provides sufficient examples to make appropriate custom or model updates how to affect accuracy. It also relies on data scientists who may lack the business environment, so it is difficult to completely capture organizational logic.
In addition, RAG is used only for fine -tuning instead of actual use, which consumes about 80 % of AI infrastructure and token, which has attracted ROI’s attention. It also lacks built-in governance-compliance teams cannot verify the appropriateness of training or ensure appropriate access control.
Illumex’s generic semantic structure (GSF) solves these challenges by automatically constructing without consumption of external AI token cards. It eliminates the needs of professional data scientists, and provides complete transparency mapping and reasoning through Web, Slack or Teams interfaces. GSF includes the clear indicators of built -in governance and interpretability, the clear indicators of tissue coverage and data quality, and the automation quality assessment of question capabilities.
Despite a large amount of investment in data infrastructure, many companies are still trying to make data -oriented decisions. Why do you think there is such a gap and how do Illumex solve it?
The gap between data investment and effective decisions continues to expand with the amount of data internally and outside explosions. Now, the organization must not only face its own growing data, but also face a series of external resources-from the weather API to the industry cloud platform, share medical care data from European institutions, and synthetic data of various use cases.
The challenge is that the organization still relies on humans to complete key data tasks such as modeling, quality assessment and instrument board creation. However, the scale and complexity of the modern data environment make the human team more and more impossible to classify data, evaluate its quality and ensure that it is suitable for AI -driven analysis and automation.
Illumex has brought about this gap through the traditional manual process of automation, enabling the organization to effectively manage and verify and use its expanded data pattern to achieve meaningful business decisions.
With the fastest industry in the Illumex platform, what unique challenges or opportunities did you observe in these fields?
We see the fastest adoption in industries where data intensity and strict supervision are used. The company needs strong data quality monitoring, follow -up and conflict detection. Financial services, drugs and retail/e -commerce are leading the charges, because these departments aim to quickly use existing data assets to quickly reshape themselves while browsing complex regulatory requirements.
With the rapid development of the generated AI, what suggestions do you provide to companies that are effective and responsible for the hope of hope?
First formulate a clear strategic plan that determines specific use cases and promotes the business command used by AI. It is important to avoid creating a new AI technology island. This technology isolate from existing systems.
Instead, establish a unified platform for integrating data management, analyzing and generating AI functions. Keeping AI initiative with the established governance practice will not only bring major risks, but also cause costs. The key is to create a shared infrastructure that supports all these functions while maintaining appropriate supervision.
With the acceleration adopted by AI, what do you think will you see in the next 3-5 years to shape the AI landscape?
There are two main trends in AI landscapes. First of all, agency analysis is obtaining motivation, so that more complicated data analysis and insight can be performed. Secondly, we see the transformation of an agent, which can achieve the workflow based on the collaboration between multiple AI models with different functions.
This arrangement allows us to surpass a single application and can achieve more comprehensive solutions. For example, in medical care, instead of isolation applications for specific tasks, consider the automation of the entire physician office workflow-scan image scanning, prescription treatment and drug recommendation in a seamless system.
These progress depends on a powerful genetic semantic structure to ensure accurate data access between AI agents, sharing context and coordination. This foundation is essential for AI solutions to achieve agency analysis and carefully planned.
Thank you for your outstanding interview, and hope that more readers should visit Illumex.
