Divel - Digital Evidence Locker

Divel secures evidence from capture to courtroom with blockchain proof, cloud storage, and AI intelligence which makes justice transparent, verifiable, and trustworthy.

Comment

Divel — Digital Evidence Locker

Inspiration

While exploring how digital evidence is used in real investigations, we noticed a recurring issue. Evidence like CCTV footage, mobile recordings, and digital documents plays a critical role in solving cases, yet its authenticity is often questioned. Digital files can be edited, duplicated, or mishandled during transfer. By the time they reach court, proving that they are original becomes difficult. This creates delays, weakens cases, and reduces trust in the system. This led us to a simple question:

Can digital evidence prove its own integrity without relying on trust?

That question became the foundation of Divel — a system designed to make evidence tamper-proof, traceable, and verifiable throughout its lifecycle.

What We Learned

Building Divel pushed us to go beyond implementation and think carefully about how real systems work.

Cryptography

We used SHA-256 hashing to generate a unique fingerprint for every file:

$$ \text{Hash} = \text{SHA-256}(\text{Evidence}) $$

This helped us understand how even the smallest change in a file results in a completely different hash, making tampering detectable.

Blockchain

We learned how blockchain can act as a layer of trust by maintaining:

  • Immutable records
  • Time-stamped events
  • A verifiable history

It showed us how systems can move from relying on trust to relying on proof.

Cloud Computing

Using AWS S3 helped us understand:

  • How to store large files reliably
  • How to scale with increasing data
  • How to manage access and security

AI Integration

We explored how AI can support investigations by:

  • Summarizing large evidence files
  • Extracting key information
  • Making data easier to understand

At the same time, we realized that AI should support decisions, not replace them.

How We Built Divel

We designed Divel as a system where each component has a clear responsibility.

Evidence Upload

An authorized user uploads evidence, and the system immediately generates a SHA-256 hash. This acts as a digital fingerprint of the file.

Secure Storage

The original file is stored in AWS S3 with encryption and controlled access. This ensures the file is safely stored and available when needed.

Blockchain Recording

Instead of storing the entire file on the blockchain, we store:

  • The hash
  • Timestamp
  • User details
  • Type of action

This creates a permanent and tamper-proof chain of custody.

Verification

To verify evidence:

  1. The file is hashed again
  2. The new hash is compared with the blockchain record

$$ \text{If } H_{new} = H_{blockchain}, \text{ the evidence is authentic} $$

AI Assistance

Once integrity is confirmed, AI is used to:

  • Summarize content
  • Extract entities
  • Provide insights

The original evidence is never modified.

Challenges We Faced

Balancing Blockchain and Scalability

At first, we considered storing everything on the blockchain, but this was not practical.

Solution: Store only hashes on the blockchain and keep files in cloud storage.

Managing Security Without Overcomplication

Adding too many layers of security can make systems difficult to use.

Solution: We focused on clear role-based access and simple, secure workflows.

Handling AI Reliability

AI can sometimes produce incorrect interpretations.

Solution: We treat AI outputs as assistive and ensure that blockchain remains the source of truth.

Building for Real Use

Many systems work well in theory but are hard to deploy.

Solution: We focused on scalability, cost efficiency, and practical usability.

Key Insight

One of the most important things we learned is:

Cloud provides availability,
Blockchain provides integrity,
AI provides usability.

Divel works because each component is used for a specific purpose.

Conclusion

Divel transforms digital evidence from something that needs to be trusted into something that can be verified.

Evidence cannot be silently altered, every action is recorded, and authenticity can be independently checked.

Instead of relying on trust, the system provides transparency, accountability, and cryptographic proof.

Final Thought

Divel does not ask institutions to trust evidence. It gives them a way to verify it.

Share this project:

Updates