Criminals always try to be a step ahead of the cops. They hide, gang up, and use several tricks to escape the watchful eyes of the cops. Now here comes Crime GPT, an exclusive Large Language Model (LLM) built for the Uttar Pradesh police that can give contextual answers on criminals in a jiffy, making it easier to track and apprehend them.

Crime GPT has been fed with data on nine lakh criminals registered in the UP police records, so cops can get information on them in a jiffy. The Crime GPT is integrated with the databases of crimes and criminals in various law-enforcement agencies in the State.

“The solution also comes with features like facial and voice identification features,” Atul Rai, Chief Executive Officer and Co-Founder of Staqu Technologies, told businessline.

While present technology solutions allow the police to do simple searches using the crime database, Crime GPT can do a lot more as it is built on large datasets of records related to crimes and criminals.

The model is built within the secure servers of the UP police, trained on Staqu Technologies proprietary LLM, and hosted in the data centre of the UP police. 

“You can ask it for the background of the criminal, and the solution will map his complete antecedents by pulling all the relevant data,” said Rai.

In the background, the tool quickly sifts through all the digital data pooled together from the CCTV feed, images, audio, and text data stored in the files.

“We have witnessed a huge impact in the last few years after the implementation of the Trinetra project in parts of UP districts. Crime GPT will help in the faster retrieval of information. It will help to collate data across various law enforcement centres, helping in faster resolution,” Prashant Kumar, Director General of Police (DGP), Uttar Pradesh Police, said.

The Gurugram-based AI start-up Staqui is nine years old and has two patents to its credit. It has raised ₹15 crore in seed and pre-Series A rounds and has 125 employees.

After deploying the solution in Uttar Pradesh, the start-up said it was planning to talk to police departments in Punjab, Telangana, Rajasthan and Uttarakhand to replicate the project.

comment COMMENT NOW