Blocktrace, an Austin-based startup, is hoping to streamline blockchain analysis using artificial intelligence (AI). Sorting through thousands of blockchain addresses and millions of transactions can be time-consuming and challenging, but Blocktrace aims to make the process faster and easier by using AI to uncover trends and anomalies. The company launched in 2018, founded by software engineer Shaun MaGruder, who previously worked as the head of training at Chainalysis, a blockchain forensics firm.

Blocktrace developed a chatbot named Robby the Robot to interact with data on the Bitcoin blockchain. According to MaGruder, they chose Bitcoin because of its large sample size and history. Blocktrace stores a copy of the Bitcoin blockchain data in its database, and with the addition of an AI layer, users can ask natural language questions like a virtual assistant. The system saves time compared to the manual process, which could take up to two hours using a blockchain explorer.

Blocktrace aims to enable investigators and users to quickly find Bitcoin addresses and identify transactions with greater accuracy and detail using OpenAI’s technology. MaGruder stated that instead of requiring a data engineer or data scientist to translate natural language questions into SQL queries, Robby has already been trained on the data model and can quickly fetch results for the user. This process is more efficient and does not require continuous training for new data scientists.

While Robby is still in beta, the chatbot is expected to be made available to the public later this year. MaGruder stated that before releasing Robby to the public, the chatbot would first be tested by a closed group to ensure its stability. The company plans to gather feedback using a thumbs up and thumbs down button to determine whether the chatbot is correct or incorrect.

Blocktrace is not the only Web3 company using AI for blockchain analysis. Others using the technology include Elliptic, Chainalysis, CipherTrace, and Nansen. Andrew Thurman, a Nansen engineer, believes that algorithmic labeling and AI labeling have a significant future in blockchain analysis. Thurman added that even if there are no explicit connections, you can infer within a good level of certainty that all those wallets are tied to the same entity.