Organizations are eager to generate tangible value and ROI from generative AI. However, recent research from the Boston Consulting Group reveals that only 26% of companies have succeeded in advancing their AI use cases beyond the proof-of-concept stage to generate real value.
For organizations already using AI, many face challenges regarding the quality and reliability of large language models (LLMs). While these foundation models can effectively offer natural language experiences, they’ve also been trained on public data, which can result in inaccuracies— especially when LLMs interact with enterprise relational databases, because they tend to struggle with the information’s complexity, structure and domain-specific nature.
To help resolve these issues, Gigaspaces designed eRag to bridge the gap between LLMs and enterprise databases. eRAG makes structured data comprehensible to LLMs, which enables companies to gain value from their enterprise datasets. To enhance the solution’s capabilities, Gigaspaces collaborated with IBM to embed watsonx.ai, IBM’s enterprise-grade AI developer studio, to deliver an end-to-end SaaS solution. This means eRAG customers can access a range of foundation models without the integration efforts typically required.
Designed specifically for structured enterprise data, eRAG’s powerful semantic reasoning engine allows LLMs to understand the meaning of operational structured data and provides relevant context needed for accurate responses. This helps non-technical users to simply and easily converse with multiple relational databases ad hoc, using natural language, to receive the correct information they need to quickly complete everyday tasks, such as customer support and inventory management.
“eRAG’s power lies in the promise of simplicity and spontaneity for end users, combined with speedy ROI and easy SaaS based onboarding for IT teams. Users can ask ad hoc questions in their own words, and get a quick, clear and accurate answer pulled from multiple complex data sets,” said Adi Paz, GigaSpaces CEO. “Our collaboration with IBM is a win-win for customers who benefit from a ready-to-use Gen AI application that has watsonx.ai seamlessly embedded.”
Additionally, to make Gen AI interactions with enterprise databases safe, eRAG integrates with IBM watsonx.governance, IBM’s end-to-end responsible AI toolkit. With watsonx.governance, eRAG offers real-time risk monitoring and mitigation, focusing on high-risk scenarios that can arise when customers interact with proprietary enterprise data. For example, eRAG continuously monitors interactions with LLMs to help companies comply with regulations and organizational policies. Then, eRAG delivers a risk-score to watsonx.governance, allowing for early detection of governance issues.
The success of GigaSpaces eRAG is a result of innovation and partnership with IBM. It’s an example of an end-to-end SaaS application that leverages all the benefits of IBM watsonx.ai to deliver tangible business benefits to organizations. In addition, the ability to integrate externally with IBM watsonx.governance helps clients pursue their Gen AI journeys with safety and confidence. Together with IBM, we’re not just meeting industry challenges, we are driving impactful results for our customers.
Get more information about eRAG at www.gigaspaces.com
