Data integration platform provider Nexla Inc. today announced an update to its Nexla Integration Platform that expands no-code generation, retrieval-augmented generation or RAG pipeline engineering, and data governance capabilities to make enterprise-grade generative artificial intelligence accessible to everyone.

The Nexla Integration Platform is a self-service data platform that enables real-time, bidirectional data integrations and data productization. The platform automates data transformation and validation, ensuring consistent data quality and simplifying complex data pipelines.

Nexla uses AI to connect to and extract metadata and transform source data into human-readable data products called Nexsets that enable true data reuse and governance. The company’s RAG framework lets companies implement RAG for agents and assistants without coding and uses large language models during each stage to improve accuracy.

This new release introduces several new capabilities that, combined, are designed to help companies deliver enterprise-grade generative AI without having to rely on specialized AI developers.

The first new feature, Agentic Retrieval, enhances the RAG process by using an LLM to dynamically search across multiple data products rather than pulling context from just a single source like a vector database. By doing so, the approach expands the scope of relevant data while automatically enforcing security policies and improving accuracy without the need to load all data into one system.

Agentic Retrieval ensures that only the most relevant information is used during AI inferencing by intelligently ranking, filtering and combining context from various sources, helping companies build more reliable and scalable generative AI applications without requiring complex data engineering.

The next new feature, Nexla Orchestrated Versatile Agents, offers an agentic interface that allows users to build complete data pipelines using natural language commands. NOVA uses an LLM to automatically generate the necessary tasks and code, like Python and SQL, to query data sources, apply transformations and orchestrate workflows.

Users can choose to work entirely within NOVA, switch to a no-code point-and-click interface, or use pro-code mode for more advanced customization. The flexibility offered is to make it easy for teams with varying skill levels to collaborate and create complex data integrations without needing specialized engineering resources.

Next up, Nexla has announced a new Converged Integration feature that simplifies and accelerates no-code data integration. It offers prebuilt flow templates and patterns for common use cases such as extract/load/transform, direct data transfers and modular RAG pipelines. The feature allows users to quickly set up and manage diverse data workflows without writing any code.

The last announcement today is the introduction of the Nexla Data Product Marketplace, a marketplace that allows data product producers to create, publish and govern the full variety of data as fully managed data products. Data products allow users to apply data policies and enforce security at runtime. The Marketplace provides a simple way to discover data while centrally controlling all access and usage.

“While there was significant hype around generative AI in 2024, 2025 is the year for AI to solve business problems,” said founder and Chief Executive Saket Saurabh. “Nexla’s latest release enables companies to be successful with agentic AI now, using their existing resources and skill sets, by combining an enterprise-grade integration and agentic AI foundation with no-code data engineering, all powered by AI.”

Image: SiliconANGLE/Ideogram