Komprise automates dataset-to-AI process workflows

Komprise has a Smart Data Workflow Manager which it says will enable users to build workflows connecting datasets with AI services.

The VC-backed company specializes in data management and orchestration, with software that tracks unstructured data usage and tiers it to an appropriate storage tier on-premises and in the public cloud, depending on its access frequency. As part of its software capabilities, it includes data moving and data identification.

AI applications such as large language models (LLMs) respond to users’ conversational questions and requests with output, reports, code, etc. To do this they need access to data sets. These sets need maintaining and refining to be kept up to date and valid, and keep sensitive information masked or hidden where needed. Komprise’s metadata can be used to tag and track AI preparedness.

Komprise identifies two major issues: Efficiently discovering and feeding the right data to an AI platform, and enriching data sets for AI processing. It says: “Both processes are highly manual, laborious tasks that are error-prone and require meticulous data governance.” Its Smart Data Workflow Manager can fix these problems, it says.

Kumar Goswami, Komprise co-founder and CEO, said: “Leveraging AI responsibly and efficiently is a priority. We are targeting common use cases that many of our customers have brought to us as a first step and we will continue to expand the Smart Data Workflow ecosystem to encompass any AI service.”

Komprise customer Rob Behary, who is the head of systems and scholarly communications in the Gumberg Library at Duquesne University, said: “With Komprise, our librarians can process more images than ever and at faster speeds by leveraging AI to systematically tag all our digital collections.”

The Smart Data Workflow Manager provides:

  • Data Workflow Wizard – Intuitive point-and-click UI wizard to set up an AI data workflow, from searching for the right data set, configuring and tuning the AI service, defining the tags and defining how frequently the workflow should run;
  • Global Search and Analytics – Use Komprise Deep Analytics to search across an entire data estate, on-premises and in the cloud, and define the precise data set needed to feed a compute-intensive AI app;
  • Automated Workflows – Automatically run workflows and repeat the process as new data becomes available, saving time and effort and ensuring continuous data enrichment;
  • A catalog of pre-built integrations – Use with Azure, AWS and other AI services for personal identifiable information (PII) data detection, Chatbot augmentation and image detection;
  • Monitor hundreds of workflows from a single interface – See their status, how many files were processed, the runtime, the next scheduled run and any actionable errors;
  • Enrich data with tags in the Komprise Global File Index – The tags file characteristics you can query and take actions on and are stored in the Global File Index, not changing file attributes in any way;
  • Data governance and auditing with Komprise’s software maintaining logs of details such as what data is fed, when, and to which service.

Komprise Smart Data Workflow Manager is available today as an early access program to customers. It is included in the Komprise Intelligent Data Management platform. Learn more here and attend a 45-minute webinar on June 6 at 15:00 UTC, registering here.