eeceb88957
Updated the README to reflect new branding and features of TrustGraph, emphasizing its capabilities in eliminating AI hallucinations and managing private knowledge bases.
7.8 KiB
Eliminate AI Hallucinations with Private Knowledge Bases
Build accurate, trustworthy AI agents powered by your own private data. TrustGraph connects your documents, databases, tools, and agents into a unified knowledge graph with precise retrieval, full observability, and deploy-anywhere control.
Table of Contents
Key Features
- Unify your Data for Smarter AI
- Ship Intelligent Agents Powered with Your Data
- Connect Your Agents with MCP
- Run Anywhere
- Observe Your Data
- Serve Models for Private LLM Inference
- Create Many Workflows
- Control Data Access
Why TrustGraph?
Ingests structured and unstructured data (PDFs, databases, OCR, custom schemas, and ontologies) into a single system. Define tools that can search your private knowledge bases and retrieve deep relationships to eliminate hallucinations from agent responses. Connect agents with the Model Context Protocol (MCP) to leverage external tools, services, and custom workflow. Deploy all of the services, datastores, and agents locally, on-premises, or in cloud. Visualize deep data relationships with 3D GraphViz and a full observability and telemetry stack. Deploy LLMs on your hardware for full control of your data. Flows and Flow Classes enable unique agent workflows. Manage user and agent access to data with collections and knowledge cores.
Agentic MCP Demo
Getting Started
Watch TrustGraph 101
Configuration Builder
The Configuration Builder assembles all of the selected components and builds them into a deployable package. It has 4 sections:
- Version: Select the version of TrustGraph you'd like to deploy
- Component Selection: Choose from the available deployment platforms, LLMs, graph store, VectorDB, chunking algorithm, chunking parameters, and LLM parameters
- Customization: Enable OCR pipelines and custom embeddings models
- Finish Deployment: Download the launch
YAMLfiles with deployment instructions
Workbench
The Workbench is a UI that provides tools for interacting with all major features of the platform. The Workbench is enabled by default in the Configuration Builder and is available at port 8888 on deployment. The Workbench has the following capabilities:
- Agentic, GraphRAG and LLM Chat: Chat interface for agentic flows, GraphRAG queries, or directly interfacing with a LLM
- Semantic Discovery: Analyze semantic relationships with vector search, knowledge graph relationships, and 3D graph visualization
- Data Management: Load data into the Librarian for processing, create and upload Knowledge Packages
- Flow Management: Create and delete processing flow patterns
- Prompt Management: Edit all LLM prompts used in the platform during runtime
- Agent Tools: Define tools used by the Agent Flow including MCP tools
- MCP Tools: Connect to MCP servers
Knowledge Cores
A challenge facing RAG architectures is the ability to quickly reuse and remove datasets from pipelines. TrustGraph stores the results of the data ingestion process in reusable Knowledge Cores. Knowledge cores can be loaded and removed during runtime. Some sample knowledge cores are here.
A Knowledge Core has two components:
- Knowledge graph triples
- Vector embeddings mapped to the knowledge graph
Integrations
TrustGraph provides maximum flexibility to avoid vendor lock-in.
LLM APIs
- Anthropic
- AWS Bedrock
- AzureAI
- AzureOpenAI
- Cohere
- Google AI Studio
- Google VertexAI
- Mistral
- OpenAI
LLM Orchestration
- LM Studio
- Llamafiles
- Ollama
- TGI
- vLLM
VectorDBs
- Qdrant (default)
- Pinecone
- Milvus
Graph Storage
- Apache Cassandra (default)
- Neo4j
- Memgraph
- FalkorDB
Observability
- Prometheus
- Grafana
Control Plane
- Apache Pulsar
Clouds
- AWS
- Azure
- Google Cloud
- OVHcloud
- Scaleway
Observability & Telemetry
Once the platform is running, access the Grafana dashboard at:
http://localhost:3000
Default credentials are:
user: admin
password: admin
The default Grafana dashboard tracks the following:
Telemetry
- LLM Latency
- Error Rate
- Service Request Rates
- Queue Backlogs
- Chunking Histogram
- Error Source by Service
- Rate Limit Events
- CPU usage by Service
- Memory usage by Service
- Models Deployed
- Token Throughput (Tokens/second)
- Cost Throughput (Cost/second)
Contributing
License
TrustGraph is licensed under Apache 2.0.
Copyright 2024-2025 TrustGraph
Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.