Building Enterprise Graph & AI Systems

Hi, I'm Sonny Honey.
I build Knowledge Graphs & GraphRAG AI.

Neo4j Certified Graph Data Science Engineer. Specializing in connecting complex enterprise unstructured data, graph algorithms, and LLM reasoning pipelines to eliminate hallucinations.

⚡ Neo4j Certified Professional ⚡ Graph Data Science (GDS) ⚡ Universal GraphRAG

Featured Work View Resume GitHub LinkedIn

About Me

Engineering Connected Intelligence

I bridge the gap between complex network theory and enterprise AI. Relational databases break down when data becomes heavily interconnected, and standard vector search fails at multi-hop reasoning.

I specialize in Neo4j Graph Data Science (GDS), knowledge graph extraction pipelines, and hybrid GraphRAG architectures that provide LLMs with structured, explainable ground truth.

Toolkit

Graph & AI Tech Stack

Neo4j & Cypher

Aura, GDS, APOC

Graph ML

NetworkX, PyVis, Louvain

LLMs & GraphRAG

Gemini, OpenAI, Vision OCR

Python Backend

FastAPI, Pandas, Streamlit

Data Pipelines

Multi-format ETL, Chunking

Cloud & Deploy

Vercel, Streamlit, Git

Flagship Systems

Enterprise Graph AI & GDS Projects

Live interactive platforms powered by Neo4j & Google Gemini

Neo4j Aura GDS Google Gemini Flash GraphRAG PyVis Physics

Universal GraphAI Platform

An end-to-end intelligence system that ingests unstructured enterprise dossiers (PDF, DOCX, CSV, TXT, MD), builds dynamic property graphs in Neo4j Aura Cloud, and performs hybrid 2-hop GraphRAG reasoning with Google Gemini to guarantee zero hallucinations and deterministic audit trails.

5+ Formats

PDF, DOCX, CSV, TXT, MD

100% Grounded

Deterministic Audit Trail

In-Browser

Interactive Subgraph Physics

Launch Live App Request Code Access

Streamlit Cloud Notice: If waking from sleep, click "Wake Up" & allow ~15s to boot.

2-Hop Cypher Traversal Neo4j Aura Cloud

// Topological Retrieval Query

MATCH (a)-[r]->(b)
OPTIONAL MATCH (b)-[r2]->(c)
WHERE size(matches) > 0
RETURN a, r, b, r2, c LIMIT 40

Factual grounding delivered to Gemini

Neo4j GDS & Cypher Multi-Objective Routing Google Gemini Flash PyVis Physics

GlobalRoute-GraphAI: Multi-Modal Logistics Optimizer

An enterprise supply chain and maritime routing engine modeling 24 global seaports, inland intermodal rail hubs, and critical maritime chokepoints. Dynamically solves multi-objective path traversals balancing lead time, freight cost ($/TEU), and Scope 3 ESG emissions (kg CO₂) with a 4-scenario geopolitical disruption simulator (Suez/Red Sea, Strait of Hormuz, Panama Canal, and Strait of Malacca).

24 Hubs

Global Seaports & Rail Terminals

4 Scenarios

Real-Time Chokepoint Rerouting

Scope 3 ESG

Auditable CO₂/TEU Tracking

Launch Live App Request Code Access

Streamlit Cloud Notice: If waking from sleep, click "Wake Up" & allow ~15s to boot.

Multi-Objective Path Traversal Neo4j Aura Cloud

// Dynamic Corridors with reduce() Aggregation

MATCH p = (src:Location)-[r:CORRIDOR*1..6]->(dst:Location)
WHERE NONE(n IN nodes(p) WHERE n.code IN $blocked)
RETURN reduce(t=0, rel IN rels(p) | t + rel.transit_time_hrs),
reduce(c=0, rel IN rels(p) | c + rel.cost_usd_per_teu)

Dynamic rerouting via Cape of Good Hope

What I Offer

Graph Data Science & AI Services

Available for contract engineering & full-time technical roles

Knowledge Graph Modeling

Transforming siloed relational SQL databases and unstructured documents into high-performance Neo4j property graph schemas and robust ETL pipelines.

GraphRAG & AI Systems

Connecting LLMs (Gemini, OpenAI, Claude) with Knowledge Graph topologies to execute complex multi-hop reasoning with zero hallucinations and verifiable citations.

Graph ML & Analytics

Applying Neo4j GDS algorithms (Louvain, PageRank, FastRP node embeddings) for community detection, fraud ring identification, and recommendation engines.

Cypher Query Tuning

Profiling Cypher execution plans, building index-backed traversals, and optimizing sub-graph queries for high-throughput enterprise performance.

Thought Leadership

Graph & AI Engineering Insights

Read insights on LinkedIn

GraphRAG • Architectural Deep Dive

Why Vector Search Fails at Multi-Hop Reasoning: Building an Auditable GraphRAG Engine with Neo4j and Gemini

Exploring how combining Knowledge Graphs with LLMs eliminates hallucinations, unlocks multi-hop inference, and provides verifiable audit trails.

Read Full Article

Graph Data Science • Supply Chain Optimization

Beyond Shortest Path: Building an Enterprise Multi-Modal Logistics & Chokepoint Engine with Neo4j & Gemini

How Graph Data Science models maritime corridors, simulates geopolitical crises (Suez, Hormuz, Panama), and optimizes multi-objective freight trade-offs.

Read Full Article

Graph ML • Algorithms

Detecting Fraud Rings with FastRP Node Embeddings & Louvain Community Clustering

Walkthrough of projecting transaction graphs into Neo4j GDS and running unsupervised network anomaly detection algorithms.

Verified Credentials

Professional Certifications

NEO4J

Neo4j Certified Professional

Cypher query tuning, Graph Data Modeling, Subgraphs & APOC.

Issued by Neo4j

NEO4J GDS

Graph Data Science (GDS)

Graph algorithms, Louvain, FastRP node embeddings, ML pipelines.