Origin

GraphWizard originated in the ACT-IAC 2026 AI Hackathon. We needed graph algorithms for the Integrity MVP, but the evaluated problem did not require a 100GB graph service or another container to operate. GraphWizard keeps the capability in process: import the algorithms, run them against standard Go graph interfaces, and inspect the results directly.

Implementation Evidence

The public record includes MIT-licensed source, clean-room implementations from academic papers, repeatable tests, 97.3% coverage, and documentation. AI accelerated the work; maintainers review and answer for each release. GraphWizard was built for a hackathon system, not a federal production deployment.

40+ Algorithms, One Consistent API

The core API covers centrality, community detection, connectivity, embeddings, flow, matching, paths, similarity, structure, and traversal. Custom implementations trace back to academic papers; wrappers make gonum algorithms available through the same style of API.

The project has grown since the hackathon. The current repository includes in-memory and disk-backed graph representations plus database loading and streaming tools. Its Go module lists gonum, DuckDB, and SQLite as direct dependencies. They run in process when used; GraphWizard still does not require a separate graph server.

The public repository (opens in a new window)is the live source for packages, algorithms, dependencies, tests, and release history.

Design Philosophy

  • One import per domain: centrality.Betweenness(g), community.Louvain(g)
  • Consistent return types: map[int64]float64 for centrality scores, [][]int64 for components
  • Standard gonum interfaces: Works with any graph.Graph implementation
  • Cleanroom implementations: Custom algorithms follow the original academic research rather than ported code
  • 97.3% test coverage: 5 packages at 100% coverage

By the Numbers

40+
Algorithms
36
Runnable Examples
97.3%
Reported Coverage
MIT
License

Quick Example

package main

import (
    "fmt"
    "github.com/intelligrit/graphwizard/centrality"
    "github.com/intelligrit/graphwizard/connectivity"
    "gonum.org/v1/gonum/graph/simple"
)

func main() {
    g := simple.NewUndirectedGraph()
    // ... add edges ...

    // Find structural weak points
    bridges := connectivity.Bridges(g)
    fmt.Println("Bridges:", bridges)

    // Rank node importance
    scores := centrality.Betweenness(g)
    for node, score := range scores {
        fmt.Printf("Node %d: %.4f\n", node, score)
    }
}

Links

View on GitHubpkg.go.dev Docs

See It in Action

GraphWizard supplies graph analysis for provider-risk prioritization in the Integrity case study. Read how we used centrality, community detection, and connectivity algorithms to surface anomalous patterns in federal payment data.

Read the Case StudyLet's Talk