Data Management Glossary
AI Agents
AI agents are software programs that autonomously perform tasks using artificial intelligence. Also referred to as agentic AI, AI agents should be able to perceive their environment, have built-in logic to make decisions, and take appropriate actions to achieve specific outcomes. Examples of AI agents include:
- Customer service bots
- Autonomous data analysis tools
- AI copilots for IT operations
- Agents orchestrating workflows in tools like Microsoft Copilot, OpenAI, or Databricks
The 2025 Trends – Artificial Intelligence (AI) report by Mary Meeker, Jay Simons, Daegwon Chae, and Alexander Krey noted:
AI is changing how we interact with the world around us. With affordable satellite connectivity expanding access to remote and underserved regions, the next wave of internet users will likely come online through AI-native experiences – skipping traditional app ecosystems and jumping straight into conversational, multimodal agents.
The report goes on to note that, “platform incumbents and emerging challengers are racing to build and deploy the next layers of AI infrastructure: agentic interfaces, enterprise copilots, real-world autonomous systems, and sovereign models.”
AI Agents Need Unstructured Data
Enterprise AI agents are increasingly being designed to interact with unstructured data, which includes files like documents, emails, images, videos, and logs. This is because:
- 90%+ of enterprise data is unstructured. (source IBM)
- Unstructured data contains rich context and institutional knowledge essential for training LLMs and powering generative AI.
AI agents require unstructured data to:
- Answer user questions accurately.
- Summarize or translate internal documentation.
- Extract insights from contracts, emails, or logs.
- Make informed decisions based on historical data.
But AI agents can’t use what they can’t access. That’s where the challenge lies.
Challenges in Feeding Unstructured Data to AI
Data silos across storage platforms (on-prem NAS, cloud object storage, etc.) contribute to a variety of issues:
- Massive data volumes to search from as many enterprises have petabytes of file data.
- High cost of storing and processing irrelevant or cold data.
- Data governance and security requirements.
- Lack of visibility into what data is useful for AI.
Komprise Helps Enterprise IT with AI Data Readiness
Komprise gives enterprise IT a flexible platform to manage unstructured data at scale, independently of data storage platform. Here’s how Komprise can be part of an agentic AI strategy:
Data Discovery & Curation
Komprise Deep Analytics and custom tagging are used to search, find and classify and unstructured file and object data across environments.
- Helps curate the “right data” for AI uses on based on age, type, owner, access frequency, and other characteristics
- Supports intelligent data selection for LLM training or inferencing
Intelligent Data Tiering
Komprise automatically moves cold data to cheaper storage based on a data management policy (e.g., cloud object storage) without disrupting user or application data access.
- Optimizes data storage costs while ensuring AI agents can still reference older data when needed.
Metadata Catalog & Indexing
Komprise builds a rich metadata catalog, or metadatabase, across all unstructured data, decoupled from storage.
- Integrate with AI data pipelines, enabling search, filtering, and ingestion of specific unstructured datasets.
Data Mobility
Komprise can be used to find, move, copy, etc. curated datasets to AI/ML platforms.
- Supports multi-cloud and hybrid cloud use cases.
Storage-Agnostic AI Data Foundation
Because Komprise works across data storage vendors (Pure, VAST, Dell, NetApp, AWS, Azure, Wasabi, etc.), it liberates unstructured data from vendor lock-in, which is critical for enterprises building vendor-neutral AI strategies.
AI Agents Need Unstructured Data
Here is a summary of the importance of unstructured data for AI agents and agentic AI success as well as the role Komprise Intelligent Data Management can play:
- Data sprawl: Find and feed the right data with a global file index (metadatabase), analytics and data tagging.
- Cost management: Efficient access to data locked in storage silos with intelligent data tiering and archival.
- Scalability: Handle petabyte-scale data workloads with parallelized indexing and mobility. (Learn more about the Komprise architecture.)
- Data curation: Prepare AI-ready data with a metadata catalog and smart data filtering and unstructured data classification.
- AI/ML integration: Input to LLMs and models with a data movement to feed AI platforms.
