Our latest community show featured a live look at Reltio AgentFlow™ Unstructured, presented by Maanasa Gottipati, Principal Product Manager. The session explored how organizations can turn document-heavy, unstructured content into trusted, usable master data—without relying on manual extraction or disconnected workflows.
The discussion centered on a growing enterprise challenge: valuable business context is often trapped in contracts, clinical documents, KYC forms, invoices, and other files. Reltio AgentFlow Unstructured helps bring that information into the Reltio Context Intelligence Platform™, where it can be extracted, mapped, and activated as part of a trusted system of context.
What Reltio AgentFlow Unstructured does
Maanasa walked through how the solution ingests unstructured documents and transforms them into structured data that can enrich existing records or create new ones. The product is designed to help teams bring more context into their master data environment and make that context available for downstream processes, analytics, agents, and AI-driven workflows.
Key capabilities highlighted in the session
-
AI-powered extraction of structured data from unstructured documents
-
Support for complex PDFs, including tables and nested fields
-
Human-in-the-loop review and fine-tuning during template setup
-
Automated mapping of extracted data to Reltio entities and relationships
-
Confidence scores at the attribute level for mapping verification
-
Automated pipeline creation for batch or scheduled document processing
-
Support for S3 and Google Cloud Storage at launch, with Azure Blob Storage on the roadmap
-
Initial availability tied to the 2026.1 release timeframe
Live demo walkthrough
The session included a step-by-step demo of how users can create a template from a sample document, review the extracted JSON output, refine extraction logic with natural language instructions, and verify how the data maps into the target Reltio tenant.
A supplier agreement served as the sample document. Maanasa showed how the system extracted entities such as contract, supplier, organization, product groups, discount structures, and key performance indicators—even from a complex table layout with pricing tiers and multiple product categories.
One key part of the demo was the ability to fine-tune extraction logic using natural language. For example, the team updated date formatting across the extracted output, and that logic then became part of the reusable template for future documents in the same corpus.
Why it matters
Once extracted and mapped, unstructured data becomes part of the same trusted data foundation teams already use in Reltio. That means organizations can apply the same governance, entity resolution, and downstream activation they use for master data today—now with additional context from documents and other unstructured sources.
This makes it easier to:
-
Reduce manual keying and review effort
-
Add governed, traceable context to enterprise records
-
Support stronger analytics and AI outcomes
-
Expand the usefulness of master data across more workflows
Example use cases discussed
The session also touched on sample use cases across industries:
-
Life sciences: Clinical protocols and regulatory documents
-
Financial services: KYC and onboarding documents
-
Manufacturing: Supplier agreements, manuals, and compliance documents
Questions from the audience
The Q&A focused on practical implementation topics, including crosswalk behavior, governance workflows, roadmap priorities, and cloud storage support. The team clarified that:
-
Crosswalk source selection is part of pipeline setup
-
Crosswalk IDs are currently system-generated
-
Custom fixed-value fields during ingestion are not available in the current version, but the request was noted as future input
-
Azure Blob Storage support is planned after the initial launch phase
-
The capability is expected to be included as part of AgentFlow
Looking ahead
The session closed with a forward-looking roadmap for 2026. Planned areas include:
-
Multimodal support for audio, video, and image-based content
-
RAG and knowledge search over unstructured corpora
-
More direct agent-based experiences for document interaction and automation
-
Expanded enterprise search capabilities across structured and unstructured context
Conclusion
This community show demonstrated how Reltio AgentFlow™ Unstructured can help organizations move beyond siloed documents and make unstructured content usable inside a governed, connected data foundation. By turning documents into trusted context, teams can improve operations today while laying the groundwork for more effective AI tomorrow.
#Featured
#CommunityWebinar
#AIML