Back to Arena

Glean

by Glean Technologies

System Card

OrganizationGlean Technologies
Released2019-01
Architecturegraph-rag / Enterprise Knowledge Graph + Hybrid Search
DetailsGlean builds a proprietary Enterprise Graph and Personal Graph per customer by ingesting 100+ app connectors (Slack, Salesforce, GitHub, Confluence, etc.) and enforcing real-time permissions. It combines lexical and semantic hybrid search with a self-learning language model. Agent memory is layered on top via an "Enterprise Context" foundation that tracks people, activity data, and content relationships.
Parameters—
Domainrag-retrievalknowledge-graphagent-memory
Open SourceNo
WebsiteVisit
enterprise-searchknowledge-graphhybrid-searchconnectorsagentic-aipermissions

Capability Profile

Benchmark Scores

6 of 14 benchmarks
Long-Context Retrieval
0/5
RULER
no data
NIAH
no data
LooGLE
no data
LongBench
no data
∞Bench
no data
Multi-Turn Recall
1/2
LoCoMo
6821p
MemoryBank
no data
Cross-Session Memory
1/1
LongMemEval
73.736p
Multi-Hop QA
2/3
BABILong
no data
MultiHop-RAG
71.144p
HotpotQA
75.371p
Agent Task Memory
1/1
AgentBench-Mem
71.825p
Personalization
0/1
PerLTQA
no data
Factuality / Grounding
1/1
RAGAS
75.496p
Sources:Glean vendor documentation; evaluated on HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering (Stanford / CMU, 1809)Glean vendor documentation; evaluated on MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries (HKUST, 2401)Glean vendor documentation; evaluated on LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory (Salesforce AI Research, 2410)Glean vendor documentation; evaluated on RAGAS: Automated Evaluation of Retrieval-Augmented Generation (Exploding Gradients, 2309)Glean vendor documentation; evaluated on AgentBench Memory Track (Tsinghua KEG, 2308)Glean vendor documentation; evaluated on LoCoMo: Long-Term Conversational Memory Benchmark (Snap Research, 2402)