Welcome to Tomasonjo Labs – your experimental playground for the latest and greatest in LLMs, graph-based reasoning, GraphRAG, autonomous agents, and everything Generative AI. We build bleeding-edge prototypes, tools, and benchmarks that explore the frontier of what’s possible when you combine language models with structured representations of knowledge.
We’re especially excited about:
- 🤖 LLM agents that reason, act, and collaborate
- 🕸️ Graphs as knowledge interfaces for LLMs
- 🔎 GraphRAG and advanced retrieval strategies
- 📈 Benchmarking & Evaluation for agentic systems
If you like systems that go beyond chat and actually think with structure, you’ll feel right at home.
A GenAI-powered legal assistant that uses structured extraction, graph-based representations, and LangGraph agents to analyze legal contracts.
A benchmark framework for evaluating text-to-Cypher agents using LlamaIndex. Designed to test how well LLMs can translate natural language into Cypher queries across various schemas and use cases.