Knowledge Graph
Dashboard
AI chatbot
Example chat: answer resource
Example chat: graph

Inspiration

Automotive manufacturers manage massive amounts of fragmented data across recalls, complaints, components, and suppliers. We built AI AutoGraph Investigation platform to connect these data sources and make complex vehicle investigations faster, more accessible, and evidence-driven.

What it does

AI AutoGraph Investigation platform is an LLM-powered vehicle investigation platform that enables users to investigate recalls, analyze complaints, trace component-supplier relationships, and explore connected vehicle data through natural-language questions.

How we built it

Neo4j Aura: Built a knowledge graph containing 9,232 nodes and 10,496 relationships.
Python: Developed data processing, graph retrieval, and evidence validation pipelines.
GroqCloud + GPT-OSS 120B: Integrated LLM-powered question answering with evidence-grounded responses.
Streamlit: Built an interactive AI assistant and graph visualization interface.
Guardrails: Implemented parameterized queries, a fixed query catalog, and source attribution.

Challenges we ran into

Our biggest challenges were integrating heterogeneous datasets, preventing unsupported AI responses, and managing LLM API limitations. We addressed these through data standardization, evidence-based answer validation, and modular LLM provider integration.

Accomplishments that we're proud of

One of our proudest accomplishments is the seamless transition from AI conversations to interactive knowledge graph visualizations. Users can ask questions in natural language and instantly explore the relationships behind the answers through Graph View, without running another database query or LLM call. This transforms our AI Assistant from a simple chatbot into an interactive investigation tool, making complex vehicle, repair, part, and supplier relationships easier to understand and verify.

What we learned

We learned that reliable enterprise AI requires more than a powerful model—it depends on high-quality data, transparent evidence, and secure system architecture.