DiVeL – Digital Evidence Locker

Inspiration

As cybercrime continues to grow, law enforcement agencies are handling an increasing volume of digital evidence, including images, videos, documents, and audio recordings. Traditional evidence management systems often rely on manual processes that are vulnerable to data tampering, inefficient retrieval, and incomplete chain-of-custody records. Investigators also spend significant time reviewing large amounts of evidence manually.

We built DiVeL (Digital Evidence Locker) to address these challenges by combining cloud computing, blockchain, and Generative AI into a single secure platform. Our goal is to help investigators manage digital evidence with confidence while reducing investigation time and ensuring evidence integrity.

What it does

DiVeL is a cloud-native digital evidence management platform that securely stores, verifies, and analyzes digital evidence.

Key features include:

Secure upload of digital evidence
Tamper-proof integrity verification using SHA-256 hashing and blockchain
Immutable chain of custody
AI-powered evidence summarization
Intelligent investigation assistant powered by Amazon Bedrock
Fast evidence search and filtering
Secure metadata management
Role-based access for authorized users
Scalable cloud architecture for handling large volumes of evidence

The platform enables investigators to upload evidence, automatically verify its authenticity, generate concise summaries, and query evidence using natural language, significantly improving investigation efficiency.

How we built it

Frontend

React.js
Tailwind CSS
JavaScript

The frontend provides investigators with an intuitive dashboard for uploading evidence, viewing case details, and interacting with the AI investigation assistant.

Backend

FastAPI
Python

The backend manages authentication, evidence uploads, case management, metadata processing, and communication with AWS services.

Cloud Infrastructure

Amazon S3
AWS Lambda
Amazon API Gateway
Amazon DynamoDB

Evidence files are securely stored in Amazon S3.

AWS Lambda automatically downloads uploaded evidence, computes a SHA-256 hash from the file contents, stores metadata in DynamoDB, and anchors the evidence hash on the blockchain.

API Gateway securely connects backend services with the serverless processing pipeline.

Blockchain

Blockchain is used to preserve evidence integrity.

For every uploaded file:

SHA-256 hash is generated
Hash is recorded on Ethereum
Transaction hash is stored
Future verification confirms that evidence has never been altered

This creates an immutable chain of custody for every piece of evidence.

Artificial Intelligence

Amazon Bedrock powers DiVeL's intelligent investigation capabilities.

The AI can:

Summarize lengthy documents
Describe uploaded evidence
Answer investigator questions using stored case information
Assist with evidence discovery
Generate investigation insights

This reduces manual analysis and helps investigators quickly understand large evidence collections.

Challenges we ran into

One of the biggest challenges was integrating multiple cloud services into a seamless workflow while maintaining security and performance.

Ensuring that blockchain verification occurred reliably after evidence uploads required careful coordination between Amazon S3, AWS Lambda, DynamoDB, and the blockchain network.

Another challenge was designing effective prompts for Amazon Bedrock so that AI-generated summaries remained accurate, relevant, and concise for investigative use.

Managing secure evidence storage, preserving metadata consistency, and maintaining a scalable architecture across distributed cloud services also required extensive testing and refinement.

Accomplishments that we're proud of

Built a complete cloud-native digital evidence management platform
Successfully integrated blockchain for tamper-proof evidence verification
Developed an AI-powered investigation assistant using Amazon Bedrock
Automated evidence integrity verification with SHA-256 hashing
Designed a scalable serverless architecture on AWS
Created an intuitive interface for investigators and administrators

What we learned

Through this project, we gained valuable experience in:

Cloud-native application development
Serverless architecture with AWS Lambda
Secure file storage using Amazon S3
Metadata management with DynamoDB
Blockchain integration for immutable verification
Prompt engineering with Amazon Bedrock
REST API development using FastAPI
Designing scalable and secure distributed systems

Most importantly, we learned how combining cloud technologies, blockchain, and Generative AI can significantly improve digital evidence management and support modern law enforcement.

What's next for DiVeL

Our future roadmap includes:

Automated image, video, and audio forensic analysis
AI-powered object, face, and scene detection
Deepfake and media manipulation detection
OCR-based extraction from scanned documents
Voice transcription and speaker identification
Advanced semantic evidence search
Multi-agency collaboration with secure evidence sharing
Court-ready evidence reports generated automatically
Mobile application for field officers
Integration with national forensic and judicial systems

Our long-term vision is to make DiVeL a comprehensive, AI-powered Digital Evidence Management System that helps law enforcement agencies securely manage evidence, accelerate investigations, and strengthen trust in the judicial process.