YiGraph
1. Product Overview
YiGraph is an end-to-end intelligent graph data analysis agent system designed to help users quickly gain insights into key relationships from complex data.
YiGraph can automatically extract entities and relationships from various raw data sources such as logs, documents, and tables to build structured graph data. Users only need to describe business problems in natural language, and the system will automatically plan the analysis process, complete calculations, and generate clear, interpretable, and traceable analysis reports.
Internally, large language models are responsible for understanding user intent, breaking down analysis tasks, and organizing final outputs. The core technology supporting the reliability of analysis results is AAG (Analytics-Augmented Generation). AAG treats analytical computation as a core capability, invoking graph algorithms and graph systems at key stages to complete verifiable calculations, which are then interpreted and summarized by the model.
Therefore, YiGraph is not just a conversational AI that "answers questions", but an intelligent graph analysis agent that can transform business problems into executable and reviewable analysis processes.
2. What Problems Does YiGraph Solve
In many business scenarios, the real value behind data lies in "relationships":
- Connections between people in social networks
- Account and fund flows in financial transactions
- Associations between accounts, devices, and addresses in risk control
- Trading and transmission paths between enterprises in supply chains
- Entity and semantic relationships in knowledge graphs
These can all be abstracted as "graphs" (nodes + edges). However, graph data is usually structurally complex, large-scale, and has a high barrier to entry: non-professional users find it difficult to accurately translate their business problems into "graph problems", let alone select algorithms, build graphs, run analyses, and interpret results.
YiGraph's goal is to make this simple: Allow users to complete graph analysis as if "asking questions" — starting from natural language and obtaining trustworthy analysis conclusions and reports end-to-end.
3. Product Positioning and Core Value
- Positioning: Intelligent graph analysis agent platform for complex relational data (integrated graph analysis from data to conclusions)
- Target Users: Risk control/compliance/security analysis teams, data analysts, and business decision support teams (especially scenarios requiring "relational insights")
- Core Value: Discover deep relationships in big data
- "Connect" data relationships scattered across different systems
- Quickly discover hidden paths, groups, and risk transmission routes
- Output interpretable and traceable results for easy review and implementation
4. Applicable Business Scenarios (Examples)
YiGraph can flexibly adapt to different industries and business needs, covering various complex relational data analysis scenarios, including but not limited to:
- Financial anti-money laundering and suspicious transaction analysis: Automatically build transaction networks from massive transaction flows to identify abnormal fund paths and suspicious transaction loops
- E-commerce risk control and wool party identification: Integrate multi-source data such as accounts, devices, and addresses to build graphs and discover organized fraud and associated malicious behavior
- Enterprise association and risk investigation: Build graphs through enterprise, equity, and transaction relationships to penetrate complex structures and identify potential compliance and operational risks
- Park/city event analysis: Unify access control, trajectory, and event data into graphs to restore personnel relationships and event evolution processes
- Supply chain risk analysis: Integrate enterprise and transaction data to build supply chain networks, locate hidden associated risks and transmission paths
5. Core Capabilities (AAG-Driven End-to-End Graph Analysis)
YiGraph's core capabilities come from AAG (Analytics-Augmented Generation): It combines "flexible semantic understanding" with "verifiable analysis processes" to complete end-to-end processing from natural language to structured conclusions. AAG's goal is not to let large models "analyze everything themselves", but to have them invoke professional analysis modules (such as graph algorithms, graph systems, etc.) at appropriate times and organize the results into reports that users can understand and review.
AAG mainly includes three types of key capabilities:
5.1 Knowledge-Driven Task Planning
The system first understands what the user's question "wants to solve", then breaks it down into executable analysis steps, such as:
- What data fields and relationships are needed
- What kind of graph should be built (which entities, which relationships)
- What analysis methods and parameters should be used
- How analysis results should be interpreted and presented
You don't need to understand graph algorithms; the system will translate "what I want to query" into "how to do the analysis".
5.2 Algorithm-Centric Reliable Execution
YiGraph will not let the model arbitrarily "write a piece of uncontrollable code and run it". Instead, it centers on "verifiable algorithm modules" for invocation and combination, making each analysis step:
- Reproducible (same input yields stable and consistent output)
- Traceable (know which algorithms were used and which steps were executed)
- More reliable (key calculations are completed by professional modules rather than pure text reasoning)
Rich Graph Algorithm Library
YiGraph provides 212 professional graph algorithms across 21 major categories, offering powerful algorithmic support for graph analysis, network modeling, structural mining, path computation, community detection, graph querying, and graph neural network scenarios.
| Algorithm Category | Number of Algorithms | Core Capabilities |
|---|---|---|
| Basics | 15 | Graph structure validation, basic traversal, topological sorting, isomorphism detection, planarity and tournament graph checks |
| Path | 19 | Shortest paths, multi-source paths, DAG longest paths, reachability analysis, Eulerian/Hamiltonian paths and cycles |
| Centrality | 21 | Key node identification, influence evaluation, bridge node detection, propagation capability analysis, hierarchy and importance measurement |
| Connectivity & Components | 17 | Connected components, strongly/weakly connected components, articulation points/bridges, bridging structures, network robustness analysis |
| Clustering & Community | 23 | Community detection, clustering coefficients, triangle relationships, cycle detection, modularity evaluation, graph partition quality analysis |
| Tree & Spanning Tree | 9 | Minimum/maximum/random spanning trees, tree structure analysis, lowest common ancestor, chordal graphs and treewidth estimation |
| Dominance | 2 | Dominance relationship analysis in directed graphs, control-flow graph analysis, dominance frontier computation |
| Flow & Cut | 7 | Maximum flow, augmenting paths, preflow-push, capacity scaling, Gomory-Hu trees, edge/node boundary analysis |
| Link Prediction | 11 | Missing edge prediction, similarity evaluation, neighbor relationship modeling, community-aware link prediction |
| Matching & Coloring | 10 | Optimal matching, edge cover, matching validity checks, graph coloring, resource allocation and conflict detection |
| Graph Matching | 2 | Subgraph pattern matching, structural containment analysis, tree isomorphism recognition |
| Graph Polynomials | 1 | Graph invariant computation, Tutte polynomial, combinatorial structure counting |
| Cliques & Cores | 8 | Maximum clique/weighted clique detection, k-core/k-truss analysis, independent sets, dominating sets, onion decomposition |
| Distance & Measures | 24 | Network diameter, radius, center/periphery, assortativity, efficiency, redundancy, reciprocity and overall structural metrics |
| Graph Query | 10 | Node attribute queries, k-hop neighbors, common neighbors, subgraph extraction, path queries and graph edit distance |
| Directed Acyclic Graphs, DAG | 4 | DAG acyclicity checks, antichain analysis, collider structures and v-structure identification |
| Bipartite | 1 | Bipartite graph recognition and structural validation |
| Graph Hashing | 1 | Graph structure hashing, canonical representation, Weisfeiler-Lehman graph hashing |
| Small-world | 1 | Small-world reference graph generation, random rewiring and network baseline comparison |
| Sparsifiers | 1 | Graph backbone preservation, low-density approximation, spanner-based sparsification |
| Graph Neural Networks | 25 | GCN, GraphSAGE, GAT, TransformerConv, heterogeneous graph convolution and graph representation learning |
For detailed algorithm descriptions and usage guides, please refer to the Algorithm Documentation
5.3 Task-Aware Graph Construction
YiGraph will not indiscriminately build all raw data into one large graph. It will selectively extract and construct "entities and relationships relevant to the problem" based on current task needs, avoiding interference from irrelevant structures, and organize the graph into a form more suitable for execution, thereby improving efficiency and result quality.
6. Data Support and Usage Methods
6.1 Multiple Data Input Types
YiGraph supports various common data formats, adapting to different starting points from "existing graph data" to "building graphs from scratch":
- Text/Logs: Reports, audit materials, logs, event descriptions, etc.
- Existing Relational Data/Knowledge Graphs: Existing node-edge data or graph assets
6.2 Multiple Usage Methods (Suggested Naming)
To facilitate different users, YiGraph can provide multiple interaction and control granularities (names can be adjusted according to final product positioning):
- Automatic Analysis Mode: Input question → System automatically completes graph building, analysis, and report generation
- Interactive Guidance Mode: Key steps (such as data scope, entity types, analysis objectives) can be confirmed or adjusted by users
- Expert Configuration Mode: Advanced users can configure algorithms, parameters, and processes with finer granularity
7. Product Advantages Summary
- ✅ End-to-End: From raw data to graph construction, graph analysis, and report output, completed in one pipeline
- ✅ Low Barrier: Complete complex relational analysis by asking questions in natural language
- ✅ Interpretable and Traceable: Results not only "give conclusions" but also explain "how they were obtained"
- ✅ More Reliable: AAG ensures key calculations are completed by verifiable analysis modules, reducing "seemingly reasonable but unreviewable" answers
⭐️ Welcome to Star / Follow AAG for the latest product progress and capability updates
📞 Contact Us
Welcome to connect with us through the following channels:
 | rednote |  |