Agents Need Maps, Not Bigger Context Windows

2026-07-01 · Source: Gradient Flow · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Data Science & Analytics · Depth: Advanced, short

Summary

Lovelace's Elemental platform, led by CEO Andrew Moore (former head of Google Cloud AI and dean of Carnegie Mellon's School of Computer Science), addresses the critical need for structured context in enterprise AI agents. This platform automates the construction and maintenance of "context graphs" from diverse enterprise data sources, including relational tables, PDFs, news feeds, and multimodal inputs. Unlike traditional knowledge graphs for human browsing, Elemental's graphs are designed for agents to navigate complex, time-sensitive questions beyond basic RAG or large context windows. Key features include scalable entity resolution, which identifies identical real-world entities across disparate records, and a retrieval mechanism that uses randomized algorithms to surface ranked connections from millions of nodes in under a second, ensuring auditability and provenance. Additionally, Lovelace provides YottaGraph, a continuously updated public world graph, to enrich private enterprise graphs without sharing customer data. This approach offers a robust, auditable context layer for agents, moving beyond reliance on graph databases for core storage.

Key takeaway

For AI Architects and MLOps Engineers building enterprise agents, relying solely on larger LLM context windows or basic RAG is insufficient for complex, auditable tasks. You should evaluate automated context graph platforms like Lovelace's Elemental to provide agents with structured, traceable knowledge. This approach enhances agent accuracy and trustworthiness, especially in regulated environments. Consider integrating public context graphs to enrich your internal data, ensuring comprehensive and verifiable agent responses.

Key insights

AI agents require automated, auditable context graphs for complex enterprise data navigation, not just larger context windows.

Principles

• Automated graph construction scales beyond human limits.
• Entity resolution is crucial for data consistency.
• Auditability and provenance are essential for agent trust.

Method

Elemental's process involves data ingestion, entity resolution, graph construction, and agent-facing tool access.

In practice

• Use context graphs for high-stakes financial investigations.
• Implement randomized algorithms for fast graph retrieval.
• Integrate public context graphs to enrich private data.

Topics

• AI Agents
• Context Graphs
• Entity Resolution
• Lovelace Elemental
• Data Integration
• YottaGraph

Best for: CTO, VP of Engineering/Data, Director of AI/ML, AI Engineer, MLOps Engineer, AI Architect

Related on AIssential

• I talked to Google’s former AI head about messy data — AI agents require automated, auditable context graphs for effective navigation of messy enterprise data.
• OpenText's Michael Cybala on Content Aviator, Knowledge Graphs, and AI You Can Trust — Context, not just data volume, is crucial for trustworthy and valuable enterprise AI.
• The Context Gap — "Context graphs" are a market response to the self-inflicted wound of lost process knowledge due to decades of outsourcing.
• bojieli / ai-agent-book — Agent intelligence stems from the synergy of LLMs, context management, and external tools.
• Connecting the Dots with Context Graphs — Stephen Chin, Neo4j — Context graphs integrate LLMs with knowledge graphs to provide grounded, explainable, and auditable AI-driven decision-making by connecting disparate enterprise data.
• Building Reliable Agentic AI Systems — Agentic RAG systems require robust context and harness engineering for reliability and trust in regulated environments.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Gradient Flow.