Prodigy AI Solutions Verbis Graph Reviews

Unlike standard RAG or search tools, it retrieves across entities, relationships, and domain context, helping organizations get more accurate, citation-backed answers with less ambiguity. Built for complex and regulated environments, Verbis Graph supports enterprise AI agents, advanced knowledge retrieval, and scalable deployment across cloud-native infrastructures.

Verbis Graph is a graph-based enterprise AI platform designed to help organizations turn fragmented documents, internal knowledge, and connected business information into explainable, searchable, and actionable intelligence. Rather than relying only on standard search or basic RAG, Verbis Graph retrieves across documents, entities, relationships, and domain context, making it especially valuable in complex and regulated environments.

Its architecture is built to support enterprise AI agents, advanced knowledge retrieval, and scalable deployment across cloud-native infrastructures. By connecting information across multiple sources, Verbis Graph helps organizations reduce ambiguity, improve traceability, and make better use of internal knowledge for decision-making and automation.

What features make Prodigy AI Solutions Verbis Graph stand out?

• Graph-Based Retrieval: Connects documents, entities, and relationships for more context-aware retrieval.
• Ontology-Aware Structuring: Organizes domain knowledge to reduce ambiguity and improve industry-specific understanding.
• Citation-Backed Explainability: Supports traceable outputs linked to source documents and relevant evidence.
• Cloud-Native Deployment: Available in enterprise-ready delivery models, including cloud marketplace deployment.
• Scalable Enterprise Architecture: Designed for complex knowledge environments and large-scale datasets.

https://x.com/prodaisolutions

What benefits should users evaluate in reviews?

• Improved Knowledge Accuracy: Helps users retrieve more relevant and context-aware information.
• Better Explainability and Trust: Makes outputs easier to verify through source grounding and connected evidence.
• Higher Efficiency: Reduces time spent searching across fragmented documents and systems.
• Support for Enterprise AI Use Cases: Provides a strong knowledge layer for AI agents, compliance workflows, and advanced decision support.
• Scalability for Complex Environments: Adapts to demanding enterprise and regulated use cases.

In sectors such as finance, healthcare, life sciences, compliance, legal, and government, Verbis Graph (verbisgraph.com) can support high-value knowledge workflows where explainability, traceability, and connected context matter. Organizations can use it to improve enterprise search, policy and compliance intelligence, contract and document investigation, scientific and clinical knowledge workflows, and other AI-driven applications that depend on reliable retrieval over complex knowledge environments.

HPC and Technical Validation

Verbis Graph has been evaluated on the Leonardo High Performance Computing (HPC) infrastructure at CINECA (Italy) using standard information retrieval benchmarks to assess large-scale graph construction and retrieval quality. 

• Enterprise-scale environment: Evaluations used NVIDIA A100 64 GB GPUs and a locally hosted 32-billion-parameter language model, demonstrating that Verbis Graph can operate with private, locally deployed AI infrastructure rather than relying exclusively on external model APIs.
• Multi-domain evaluation: Evaluation covered biomedical, scientific-evidence, financial and legal information-retrieval datasets.
• NFCorpus biomedical evaluation:
  • Initial retrieval hit rate: 89.16%
  • Initial MRR: 0.303
  • Average baseline retrieval latency: 1.34 seconds
  • Improved retrieval hit rate: 92.26%
  • Improved MRR: 0.578
  • Subsequent embedding configuration MRR: 0.592
  • Subsequent Hit@1: 50.5%
• SciFact scientific-evidence evaluation:
  • MRR: 0.721
  • MRR with reranking: 0.734
  • Hit@100: approximately 98%
• FiQA financial-domain evaluation:
  • MRR: 0.584
  • Retrieval coverage: 95.3%
  • Hit@1: 51.6%
  • Hit@10: 74.2%
  • Mean recall: 85.3%

Read complete report verbisgraph.com/uploads/VerbisGraph-Evaluation%20of%20Graph-Enhanced%20Retrieval%20on%20HPC%20Infrastructure.pdf