Data productivity provider Matillion announced its AI vision, with a range of GenAI functionality to put AI in the hands of every data practitioner, coders and non-coders alike.
The addition of a low-code/no-code graphical AI Prompt component will enable every data engineer to harness prompt engineering within LLM-enabled pipelines and materially boost productivity, whilst unlocking the infinite opportunities of unstructured data.
The model-agnostic design will allow users to choose their preferred LLM, set the right context, and drive prompts at speed and scale. Seamlessly integrating with existing systems, the technology will enable information extraction, summarization, text classification, NLP, sentiment analysis and judgment calls on any source that Matillion connects to.
With a strong emphasis on security and explainability, the solution safeguards data sovereignty, transparently articulates results and actively eliminates bias.
LLM-enabled pipeline functionality within Matillion is expected to launch in the first quarter of 2024.
Ciaran Dynes, Chief of Product at Matillion, said: “The role of the data engineer is evolving at pace. With the advent of GenAI, data engineering is about to get much more interesting.
“Matillion’s core ethos is to make data more productive, and enabling users to seamlessly integrate AI into their data stack and leverage that functionality without the need for a data scientist, is doing just that.
“Whilst all eyes are on AI, BI isn’t going anywhere. We believe that through the Data Productivity Cloud, we have the opportunity to democratise access to AI in the context of data pipelines to augment BI projects, and to train and consume AI models.”
Also Read: Microsoft and Octopai to Co-Sell Leading Innovative Data Management Integration
Addressing the data security concerns around GenAI uptake, Matillion is set to create a stack-ready AI-enabled solution that pushes down processing into cloud platforms such as Snowflake’s Snowpark, allowing Python to run inside Snowflake at scale while maintaining business data sovereignty.
Matillion’s GenAI feature list, to be added to its Data Productivity Cloud platform throughout the coming year, simplifies the complexities of AI adoption for data engineers. Matillion is excited to announce a planned co-pilot feature alongside highly sought-after auto-documentation capabilities.
Matillion plans to launch further AI innovation throughout 2024, including Pipeline RAG (Retrieval Augmented Generation). Matillion will allow integration with vector databases to supercharge LLMs with external customer knowledge and go beyond their built-in training data. This enables users to bring their business data to LLMs to achieve more personalised, accurate responses from AI with better governance.
Also, in 2024, Matillion plans to roll out graphical components that will empower data teams to build fine-tuning data pipelines, in order to adjust LLMs with specific task data. Leveraging the connectivity and transformation capabilities of Matillion, teams can prepare AI-ready data and fine-tune models to excel on specific tasks like never before.
Matillion’s AI vision announcement follows the launch of the Data Productivity Cloud earlier in 2023, a unified platform to build pipelines and deliver data at unlimited scale. Matillion was minted a unicorn in 2021 with a valuation of $1.5bn, and recently opened new HQs in both Manchester, UK, and Denver, CO.
SOURCE: PRNewswire