Three winners this week! 
Congratulations to @sekharchennaiah12345, @leongkwokhing, and @xlrtaha!
Also, the Round of 32 Challenge is now open, with your choice of a LEGO Cristiano Ronaldo CR7 or Lionel Messi #10 Celebration Display up for grabs.
GraphAcademy Cup: Week 3 Community Challenge Winners
Week 3 asked you to Show What You Learned, and the community delivered.
The quality of the submissions made judging especially difficult, so instead of selecting a single winner, we've decided to recognize three outstanding projects. Each of these submissions demonstrates creative thinking, solid graph modeling, and a practical application of what was learned in GraphAcademy.
Congratulations to this week's winners!
Winners will receive a direct message within the next 3–5 business days to confirm eligibility and arrange prize shipping.
KiranaAI: Text2Cypher + Text2SQL AI Agent for Kirana Stores
Congratulations @sekharchennaiah12345!
KiranaAI is an AI assistant built for local kirana stores that lets owners ask business questions in Tenglish (Telugu + English). Behind the scenes, a LangGraph ReAct agent determines whether the question should be answered with Neo4j, SQLite, or both.
What stood out:
- A well-designed graph model connecting customers, suppliers, products, referrals, and store credit
- Reliable Text2Cypher generation grounded in the live Neo4j schema
- Excellent documentation with screenshots showing the complete flow from natural language to generated Cypher and graph visualization
- A thoughtful solution built around a real-world business problem
Submission: 🔴CLOSED: Week 3 : KiranaAI (Text2Cypher + Text2SQL AI Agent for Kirana Stores - India) - 🇮🇳
GitHub: GitHub - chakka-guna-sekhar-venkata-chennaiah/kirana-ai-agent · GitHub
IT Resilience Graph
Congratulations @leongkwokhing!
This project models enterprise infrastructure in Neo4j Aura, making it easy to understand how failures propagate across servers, applications, networks, and external vendors.
What stood out:
- Practical use of graph modeling for dependency analysis
- Clear blast-radius queries using variable-length path traversal
- Vendor risk analysis that illustrates downstream business impact
- Strong use of concepts from the Graph Data Modeling Fundamentals course
Submission: 🔴 CLOSED: Week 3 Challenge: Show What You Learned - #3 by leongkwokhing
VEIL: Graph-Native Corporate Relationship Platform
Congratulations @xlrtaha!
VEIL imports data from the UK Companies House registry into Neo4j, allowing users to explore corporate ownership structures, directors, officers, and beneficial ownership through an interactive Streamlit application.
What stood out:
- Integration with a real public data source
- Clean graph model for exploring ownership relationships
- Useful traversal patterns for compliance and investigative workflows
- A strong example of where graph databases naturally outperform relational approaches
Submission: 🔴 CLOSED: Week 3 Challenge: Show What You Learned - #6 by xlrtaha
Demo: ubo-knowledge-graph | Devpost
Round of 32 Challenge Is Now Open
With the group stage complete, the knockout rounds begin.
This week's challenge is based on the Graph Data Science learning path and follows one simple rule:
The Rule of 32
- Maximum 32 nodes
- Maximum 32 relationships
- Project description between 32 and 320 words
This round runs July 2 through July 8.
Unlike Week 3, there will be one winner. The winning project will earn its creator a choice of either the LEGO Cristiano Ronaldo CR7 Celebration Display or the LEGO Lionel Messi #10 Celebration Display.
Challenge details and submission thread
If you've been working through the Graph Data Science learning path, this is a great opportunity to put those concepts into practice. We don't expect large projects. Sometimes the simplest graph tells the best story.
Good luck, and we're looking forward to seeing what you build. 