Police Automating Interview

Project description

The project:

• The stakeholders = The Dutch National Police, The Vice Squat, precisely Esther van Gloven, Intern Deputy Chief Vice Squat.
• The problem = Vice Victims repeating there story in multiple interviews.
• Example = A Vice Victim calls the police to report what happened to him/her. He/She will have to repeat their story multiple times, to multiple police departments, hospital staff, doctors, psychologists etc.
• Consequence = This could cause even more psychological trauma to the victim.
• Stakeholder’s desire = An AI solution that will minimize the repetition of the victim’s story to only 1.

For this project, our team has been tasked to create an AI solution that will minimize the repetition of the victim’s story to only 1. This will be done by creating an AI that takes the recording of the very first interview with the victim as input. This input has to be processed to text (using Speech-To-Text). After that, the text has to be logically distributed to the different departments and institutions based on what each one of them needs from the interview. The stakeholder’s intention with the work we will produce is to take it, and refine it to their liking, policies, etc. And finally, integrate it into their infrastructure, where the actual recording of the interview will find place.

Context

This project helps the Dutch National Police’s Vice Squad. Right now, vice victims have to repeat their stories many times (police, hospitals, doctors, psychologists), causing extra trauma. Our AI solution will reduce this to just one interview. We record the victim’s first interview, convert it to text, and then share the needed parts with each group. Finally, the Vice Squad will refine and use this solution in their system. This way, victims can avoid extra stress and re-traumatization.

Results

The key outcomes of this project include the development of an AI-based solution that minimizes the repetition of vice victims’ stories and improves the overall efficiency of the process. The product includes a system that records the victim’s first interview, converts it into text using Speech-to-Text technology, and distributes the necessary parts of the text to relevant departments, such as police, healthcare, and psychologists.

This solution has two main benefits:

1. Reduced Trauma for Victims: By ensuring victims only need to share their story once, we reduce the emotional burden and prevent further psychological harm.
2. Increased Efficiency: Automating the distribution of information saves time for both victims and authorities. Departments no longer need to re-interview victims for the same information, which streamlines workflows and reduces delays.

About the project group

Our team =

• Said Alhazina – Infrastructure student – semester 3
• Arif Yildirim – Infrastructure student – semester 3
• Kamen Trifonov – Infrastructure student – semester 3
• Stefan Popescu– Software student – semester 6

Stefan focused on developing the AI.
Said, Arif and Kamen focused on creating a secure, reliable and scalable infrastructure where the AI can run.
We've spent the first half of the semester on research & analysis and the other half on actual implementation.