Zero-Knowledge AI Data Vault

[Proposal setup] Proposal title

Please provide your proposal title

Zero-Knowledge AI Data Vault

[Proposal Summary] Time

Please specify how many months you expect your project to last

3

[Proposal Summary] Translation Information

Please indicate if your proposal has been auto-translated

No

Original Language

en

[Proposal Summary] Problem Statement

What is the problem you want to solve?

AI needs high-quality data, but strict privacy rules block sharing. Training is opaque, compliance is hard to verify, users lack control, and impeding collaborations among institutions.

[Proposal Summary] Supporting Documentation

Supporting links

• https://github.com/jyanokye
,
• https://github.com/ThePaapyK
,
• https://github.com/orgs/Cardano-Cart/repositories
,
• https://github.com/ThePaapyK/GalamseyWatch

[Proposal Summary] Project Dependencies

Does your project have any dependencies on other organizations, technical or otherwise?

No

Describe any dependencies or write 'No dependencies'

No dependencies

[Proposal Summary] Project Open Source

Will your project's outputs be fully open source?

Yes

Please provide here more information on the open source status of your project outputs

All code, UI, tests, documentation, and demo materials will be released under MIT License and hosted publicly

[Theme Selection] Theme

Please choose the most relevant theme and tag related to the outcomes of your proposal

AI

[Campaign Category] Category Questions

What is useful about your DApp within one of the specified industry or enterprise verticals?

DApp Overview

• Purpose: Enable privacy-preserving AI training on sensitive datasets (health, finance, research, education) while generating verifiable zero-knowledge proofs (ZKPs).
• Industry Vertical: AI (Area 9 – Responsible Privacy-Preserving AI)
• Key Features:
  • Encrypt datasets at source
  • Train AI models on encrypted data without exposing raw datasets
  • Generate ZKPs to verify policy compliance
  • Auditors can verify compliance without accessing raw data

Compact Contracts & File Paths

• src/contracts/DataVault.compact.ts → store encrypted datasets
• src/contracts/ZKTraining.compact.ts → train models & generate ZKPs
• src/api/dataBridge.ts → entrypoints: uploadEncryptedData(), trainModel(), verifyProof()
• src/ui/ → demo UI for uploading datasets, triggering training, and verifying ZKPs

Wallet Integration

• Lace (Midnight) wallet for authentication and signing

Test Commands

npm run test:encrypt      # Tests data encryption module

npm run test:trainZKP     # Tests training pipeline with ZK proofs

npm run test:audit        # Tests audit verification workflow

What exactly will you build? List the Compact contract(s) and key functions/proofs, the demo UI flow, Lace (Midnight) wallet integration, and your basic test plan.

2.1 What We Will Build

Compact Contracts

• DataVault.compact.ts

  • Functions:

    • storeEncryptedData(data: EncryptedDataset) → Store encrypted datasets securely

    • retrieveEncryptedData(datasetId: string) → Retrieve encrypted dataset for AI training

• ZKTraining.compact.ts

  • Functions:

    • trainModelZKP(datasetId: string, modelParams: ModelConfig) → Train AI model on encrypted data

    • verifyCompliance(trainingProof: ZKProof) → Generate and validate zero-knowledge proof of policy compliance

  • Proofs:

    • Zero-knowledge proofs ensure AI training adhered to permitted rules without revealing raw data

Demo UI Flow

1. User logs in via Lace (Midnight) wallet

2. Upload encrypted dataset → stored in DataVault.compact.ts

3. Trigger AI training → ZKTraining.compact.ts performs computation on encrypted data

4. Display results + ZKP verification in the UI for audit or compliance purposes

API Entrypoints

• src/api/dataBridge.ts

  • uploadEncryptedData() → Connects UI upload to DataVault.compact.ts

  • trainModel() → Triggers training on encrypted datasets

  • verifyProof() → Retrieves and verifies ZKP from ZKTraining.compact.ts

Basic Test Plan

npm run test:encrypt      # Verify encryption of datasets

npm run test:trainZKP     # Test AI training pipeline with ZKPs

npm run test:audit        # Validate audit / compliance verification workflow

How will other developers learn from and reuse your repo? Describe repo structure, README contents, docs/tutorials, test instructions, and extension points. Which developer personas benefit, and how will you gauge impact (forks, stars, issues, remixes)?

• Clear Repo Structure:

  • /src/contracts → Compact contracts for encrypted data storage and ZK training

  • /src/api → Functions connecting UI to contracts (uploadEncryptedData(), trainModel(), verifyProof())

  • /src/ui → Demo interface showing dataset upload, AI training, and ZKP verification

  • /test → Unit and integration tests demonstrating contract behavior

  • /docs → Tutorials, API references, and workflow guides

• Comprehensive Documentation:

  • Step-by-step instructions for uploading datasets, training models, and verifying ZKPs

  • Examples of API calls and contract interactions

  • Guidance on extending PoC to other AI models or datasets

• Tutorials and Examples:

  • End-to-end demo showing privacy-preserving AI workflow

  • How to generate and verify zero-knowledge proofs

  • Integration with Lace (Midnight) wallet for authentication

• Extension Points:

  • Add new AI models for encrypted training

  • Extend ZKP logic for more compliance scenarios

  • Explore additional privacy-preserving patterns using Compact

• Learning Outcomes for Developers:

  • Practical understanding of privacy-preserving AI on Midnight

  • Experience integrating Compact contracts and zero-knowledge proofs

  • Template for building reusable, privacy-first DApps

[Your Project and Solution] Solution

Please describe your proposed solution and how it addresses the problem

Problem Perception

• Organizations and researchers need access to high-quality datasets for AI training.
• Sensitive data (health, finance, research, education) cannot be shared due to privacy laws and ethical concerns.
• Lack of collaboration slows AI innovation and exposes users to privacy risks.

Proposed Approach

• Use Compact TypeScript contracts to manage encrypted datasets (DataVault.compact.ts) and AI training (ZKTraining.compact.ts).
• Employ zero-knowledge proofs (ZKPs) to ensure that AI models are trained in compliance with rules without revealing raw data.
• Integrate Lace (Midnight) wallet for authentication and signing of transactions.
• Provide a demo UI for uploading data, triggering training, and verifying proofs.

Project Engagement

• Developers: Learn privacy-preserving AI workflows using Compact contracts and ZKPs.
• AI Researchers: Access secure, encrypted datasets for collaborative training without violating privacy regulations.
• Auditors / Regulators: Verify compliance through ZKP proofs without exposing sensitive data.

Uniqueness & Benefits

• First Midnight-native PoC enabling AI model training on encrypted, sensitive datasets.
• Introduces a reusable building block for zero-knowledge AI collaboration.
• Benefits:
  • Developers: Can replicate and extend PoC for new AI use-cases.
  • Institutions: Collaborate securely on sensitive datasets.
  • Midnight ecosystem: Demonstrates practical privacy-preserving AI workflows, encouraging adoption of Compact contracts.

Importance to Midnight

• Shows how Midnight + Compact enables confidential, auditable AI workflows that are impossible on public ledgers.
• Provides an open-source reference PoC that encourages developers and enterprises to build privacy-preserving DApps on Midnight.
• Supports the ecosystem’s goal of responsible AI and data privacy at scale.

[Your Project and Solution] Impact

Please define the positive impact your project will have on Midnight ecosystem

Value to Midnight

• Demonstrates practical privacy-preserving AI workflows on Midnight using Compact contracts.

• Provides a reusable open-source PoC for developers, accelerating adoption of Midnight technology.

• Enables regulated institutions to collaborate on AI models without violating privacy laws, showcasing Midnight’s real-world utility.

• Strengthens Midnight’s positioning as a platform for responsible, auditable AI development.

Measuring Impact

• Quantitative Metrics:

  • GitHub repository activity: forks, stars, issues, pull requests

  • Number of PoC remixes or extensions by other developers

  • Test suite usage and contributions from the community

• Qualitative Metrics:

  • Developer feedback via tutorials and walkthroughs

  • Adoption stories or references from enterprise or academic partners

  • Requests for new features or privacy patterns inspired by the PoC

Sharing Outputs

• Public GitHub repository containing all Compact contracts, API, UI, tests, and documentation

• Step-by-step tutorials and README to guide developers through the PoC workflow

• Demo videos and walkthroughs to demonstrate end-to-end encrypted AI training with ZKPs

• Open channels for developer engagement, including issues, discussions, and community contributions

• Presentations at Midnight ecosystem events or hackathons to encourage adoption and collaboration

[Your Project and Solution] Capabilities & Feasibility

What is your capability to deliver your project with high levels of trust and accountability? How do you intend to validate if your approach is feasible?

Demonstrated Capability

The ZK-AIDV project team brings proven expertise in blockchain development, zero-knowledge proofs (ZKPs), AI integration, and front-end engineering. Our ability to deliver high-quality outcomes is reinforced by the successful completion of previous funded projects, such as:

• Cardano E-Commerce Application (https://projectcatalyst.io/funds/12/cardano-open-developers/cardano-ecommerce-application)
  • Delivered all milestones
  • Built a fully functional proof-of-concept (PoC)
  • Maintained full accountability for allocated funds

This prior success demonstrates our technical competence, reliability, and ability to manage funded projects effectively.

Feasibility Validation

The feasibility of ZK-AIDV will be validated through a structured, multi-step approach:

• Incremental Milestones:
  • 3-month plan with clearly defined milestones:
    1. Proof circuit & Compact smart contract development
    2. UI integration with Lace (Midnight) wallet
    3. Final documentation, test suite completion, and PoC demonstration
  • Ensures measurable progress and early detection of risks at each stage
• Test-Driven Development:
  • Rigorous testing of zero-knowledge proof generation and verification
  • Includes edge cases such as invalid or expired proofs
  • Guarantees robustness and reliability of the solution
• Proof-of-Concept Demonstration:
  • Working PoC showcasing:
    • Encrypted AI training workflow
    • ZKP verification of compliance rules
    • Wallet authentication and user interaction
  • Provides tangible evidence of technical feasibility

Trust & Accountability

Our experience managing funded projects demonstrates that we can responsibly manage resources. For ZK-AIDV, we will maintain this high standard through:

• Budget Tracking: Clear allocation of funds to each milestone and deliverable
• Regular Reporting: Transparent updates and progress reports to stakeholders
• Open-Source Commitment: Public repository with code, documentation, and setup guides under MIT license
• Version Control & Auditability: GitHub repository ensures traceable and auditable development

Combining prior experience, structured planning, transparent processes, and technical expertise positions the team to deliver ZK-AIDV reliably, on time, and with proper stewardship of funds.

[Final Pitch] Budget & Costs

Please provide a cost breakdown of the proposed work and resources

Budget Breakdown

Total Requested Amount: $10,000 USD

Labour & Subcontracting Costs

• Amount: $7,000 (~70% of total budget)
• Covers the time and effort of the core team delivering all three milestones:
  • Anokye James Yaw – Blockchain development, ZKP circuits, smart contracts, milestone oversight
  • Paa Kojo Effah Annan – Front-end development, UI integration, wallet auth, testing, and documentation
• Justification: Labour is the largest cost due to the highly technical nature of contract development, AI integration, and PoC creation.

Overhead Costs

• Amount: $1,000 (~10% of total budget)
• Covers infrastructure and operational support:
  • Computers and workstations
  • Office space and IT services
  • Software licenses (IDEs, testing tools, libraries)
• Justification: Essential to ensure the team has a functional environment to execute the project efficiently.

Materials Costs

• Amount: $1,000 (~10% of total budget)
• Covers third-party materials and services directly used in the project:
  • Cloud storage or test network usage for encrypted datasets
  • Paid libraries or ZKP generation/verification tools
• Justification: These materials are necessary for building, testing, and demonstrating the AI PoC.

Optional / Miscellaneous Costs

• Amount: $1,000 (~10% of total budget)
• May cover eligible additional items if required:
  • Hackathon prizes or bounty rewards to encourage community participation
  • Travel or venue hire for live demos or workshops
  • Training for team members on specific tools or Midnight integration
  • Licensing of third-party technology for PoC expansion
• Justification: Reserved for optional costs that enhance project impact but are not central to core delivery.

Summary

• Labour & subcontracting: $7,000
• Overhead: $1,000
• Materials: $1,000
• Optional/miscellaneous: $1,000

This allocation ensures that the majority of funds are spent directly on project delivery, with a portion allocated to infrastructure, materials, and optional enhancements, maintaining responsible and transparent use of the grant.

[Final Pitch] Value for Money

How does the cost of the project represent value for the Midnight ecosystem?

The ZK-AIDV project represents a high-value use of the $10,000 grant because it directly advances privacy-preserving AI on the Midnight platform, creating reusable building blocks for developers and accelerating the adoption of secure, compliant AI workflows.

How the Project Creates Value

1. Demonstrates Novel Capabilities: ZK-AIDV provides a working proof-of-concept for AI model training on encrypted datasets with zero-knowledge proof verification, a workflow that has not been widely implemented in other Midnight examples. This showcases the platform’s unique ability to support regulated data in a privacy-preserving manner.
2. Reusable Developer Resources: The project will produce public smart contracts, APIs, UI components, tutorials, and a fully documented PoC. This provides a ready-to-use reference for developers (DApp developers, zk-curious engineers, and integrators) to build new privacy-preserving AI applications on Midnight, encouraging broader ecosystem growth.
3. Accelerates Trust & Adoption: By demonstrating verifiable ZKP-based AI training workflows, the project helps build trust with organizations handling sensitive datasets, showing that compliance and innovation can coexist on Midnight.
4. Efficiency of Fund Use: The allocation of $10,000 ensures that the majority of funds are applied directly to labour and development of key technical milestones, while a smaller portion supports infrastructure, materials, and optional enhancements. This approach maximizes deliverables per dollar, ensuring transparency, accountability, and tangible outcomes for the ecosystem.

Measurable Impact

• Quantitative: Number of forks, stars, issues, and remixes of the public GitHub repository; usage of provided contracts and UI components in downstream projects.
• Qualitative: Developer feedback, adoption by privacy-conscious AI projects, and demonstrable PoC workflows that others can replicate.

By funding ZK-AIDV, Midnight gains a reusable, open-source, privacy-preserving AI PoC that demonstrates the platform’s strengths, encourages developer adoption, and provides a concrete foundation for future innovation in regulated AI applications.

[Self-Assessment] Self-Assessment Checklist

I confirm that the proposal clearly provides a basic prototype reference application for one of the areas of interest.

Yes

I confirm that the proposal clearly defines which part of the developer journey it improves and how it makes building on Midnight easier and more productive.

Yes

I confirm that the proposal explicitly states the chosen permissive open-source license (e.g., MIT, Apache 2.0) and commits to a public code repository.

Yes

I confirm that the team provides evidence of their technical ability and experience in creating developer tools or high-quality technical content (e.g., GitHub, portfolio).

Yes

I confirm that a plan for creating and maintaining clear, comprehensive documentation is a core part of the proposal's scope.

Yes

I confirm that the budget and timeline (3 months) are realistic for delivering the proposed tool or resource.

Yes

[Required Acknowledgements] Consent & Confirmation

I Agree

Yes