Hi everyone! I've recently become a victim of a scam from a company, and I've started doing some investigating on my own. In my research, I've uncovered a network of around two thousand suspicious companies, most of which share the same address and directors. They seem to be consistently recording losses, while one particular company in this network is generating significant profits, managed by a director from Asia or Russia. I've even developed some Python scripts with GPT to connect to various APIs and gather data about these companies. It really feels like I'm dealing with a large scam operation that's siphoning money out of the country and shutting down these hollow companies when they're no longer profitable. I'm looking for ideas on how to analyze these 2000 records for patterns and how to visualize the cash flow over the years. Are there any AI tools specifically designed for investigations like this?
2 Answers
This issue happens more often than you might think! I suggest leveraging a graph database like neo4j or Janus to help uncover patterns related to these fraudulent activities. They can simplify the process of tracking connections and detecting suspicious behavior.
You can certainly analyze your company data quite effectively! Since the records are text-based, try organizing them into JSON or a CSV format. For analyzing the data, I recommend using Gemini 2.5 Pro due to its massive 1M token context window, which should help with the quantity of records you’re dealing with. For visualizations, if financial data is structured as a CSV, you can create a Jupyter notebook to process this data and visualize fund flows using libraries like NetworkX. Plus, if you format your data as a NetworkX graph, you can analyze it further using community detection or centrality measures.
Thanks for the tip!