Mirrored from filpgf.io — ProPGF Batch 3 (Karma program 1479, application 6a274c68a253cb2d96eca598, status: pending). Contact details redacted; canonical application lives on filpgf.io. 1.1 Project Name FairPayFile: Fair Paid Exchange for Filecoin-Stored AI Data 1.4 Team Lead/Point of Contact Majid Khabbazian 1.5 Category [ "Other" ] 1.6 Open Source Status Fully Open Source 2.1 Project Summary FairPayFile will implement an open-source Filecoin-compatible prototype for fair paid exchange of large encrypted data objects, starting with AI datasets, model artifacts, scientific datasets, and other high-value data stored or retrieved through Filecoin/IPFS-compatible workflows. The project will support data owners, storage/retrieval providers, dataset-marketplace builders, and data buyers who need stronger delivery and payment guarantees than ordinary retrieval provides. The specific gap is that a buyer should not have to pay before knowing that the purchased data is correct and recoverable, while a seller should not have to release usable plaintext or decryption material before the settlement condition is satisfied. Building on our FC 2026 paper “Plaintext-Scale Fair Data Exchange,” this grant will fund implementation rather than further theory: a core library, seller/buyer CLI tools, a Filecoin/IPFS demo, a modular settlement abstraction, benchmarks, and documentation. The outcome will be reusable infrastructure for high-value Filecoin-stored data products, helping make Filecoin more useful as a substrate for paid data exchange. 2.2 Who does this work support? [ "Storage Providers", "Application Builders", "Application Users", "Network Infrastructure", "Other" ] 2.3 Total Funding Requested (USD) $95000 2.4 Milestones & Budget [ { "title": "Filecoin-Specific Architecture, Threat Model, and Implementation Plan", "description": "This milestone will translate the existing Plaintext-Scale Fair Data Exchange research prototype into a concrete Filecoin-oriented implementation plan. \nWe will define the FairPayFile architecture, including seller and buyer workflows, cryptographic metadata formats, Filecoin/IPFS-compatible storage and retrieval assumptions, and a modular settlement interface. \nWe will also produce a threat model, API design, repository structure, licensing plan, and test/benchmark plan for the implementation milestones that follow.", "dueDate": "2026-08-31", "fundingRequested": "$15,000", "completionCriteria": "Milestone 1 will be complete when the public project repository is initialized and contains: (1) a Filecoin-specific architecture document; (2) a threat model covering buyer, seller, storage/retrieval, and settlement assumptions; (3) seller-side and buyer-side API/CLI specifications; (4) a proposed cryptographic metadata format; (5) a modular settlement interface design; (6) an implementation roadmap for the remaining milestones; and (7) an initial test-vector and benchmark plan. The repository will also include the selected open-source license and contribution/readme documentation sufficient for external review." }, { "title": "Core FairPayFile Library and Seller/Buyer CLI Implementation", "description": "This milestone will implement the core FairPayFile software stack. \nWe will build the main open-source library for plaintext-scale fair data exchange, together with seller-side and buyer-side command-line tools. \nThe seller workflow will support preparing a large data object, generating the required cryptographic metadata, and producing an encrypted object suitable for storage through Filecoin/IPFS-compatible workflows. \nThe buyer workflow will support loading the metadata, verifying the exchange instance, and recovering the plaintext after the fair exchange condition is satisfied in the prototype flow. \nThis milestone will also include unit tests, deterministic test vectors, basic error handling, serialization formats, and continuous integration.", "dueDate": "2026-10-31", "fundingRequested": "$30,000", "completionCriteria": "Milestone 2 will be complete when the public repository contains a working FairPayFile core library and seller/buyer CLI tools with documented commands for preparing, verifying, and completing a local fair exchange of a large encrypted file. \nThe implementation must include deterministic test vectors, unit tests for the main cryptographic and serialization components, CI configuration, clear build instructions, and a local end-to-end demo showing that a seller can prepare an encrypted data object and a buyer can verify the metadata and recover the correct plaintext after the pro …[truncated] 3.1 Impact pathway FairPayFile’s direct outputs will be an open-source implementation of plaintext-scale fair data exchange, seller and buyer CLI tools, a Filecoin/IPFS-compatible encrypted-object demo, a modular settlement abstraction, benchmarks, documentation, and a public v0.1 release. These outputs give Filecoin ecosystem developers, storage providers, data owners, and dataset-marketplace builders a concrete mechanism for selling access to large encrypted data objects with stronger delivery/payment fairness guarantees. The immediate outcome is that high-value data exchange on Filecoin becomes easier to prototype and evaluate. A data seller can prepare an encrypted AI dataset, model artifact, scientific dataset, or licensed archive; associate it with Filecoin/IPFS-compatible storage; and let a buyer verify the exchange before the usable plaintext is released. This reduces a key trust barrier: buyers do not want to pay for incorrect or unrecoverable data, and sellers do not want to release valuable plaintext before settlement conditions are satisfied. At the network level, this can support paid Filecoin usage by making Filecoin-stored data more commercially usable. FairPayFile does not itself promise to generate paid mainnet deal volume during the grant period; instead, it provides enabling infrastructure for future paid data products, dataset marketplaces, AI-data services, and storage/retrieval provider revenue models. This indirectly contributes to driving paid onchain deals, strengthening network profitability, and scaling adoption by high-value clients who need verifiable and fair access to data. The expected change is that Filecoin becomes more than a storage backend for large objects: it becomes a stronger substrate for paid exchange of valuable data. By turning an accepted fair data exchange protocol into reusable implementation artifacts, the project lowers the cost for Filecoin builders to experiment with pay-per-file and pay-per-dataset workflows, especially in AI, scientific data, and licensed data markets. 3.2 Verification metrics N/A. FairPayFile does not directly move a specific onchain metric during the grant period. The project’s direct outputs are open-source implementation artifacts: a core fair data exchange library, seller/buyer CLI tools, a Filecoin/IPFS-compatible demo, a modular settlement abstraction, benchmarks, documentation, and a public v0.1 release. These outputs indirectly support future paid onchain deals and paid data products by reducing the trust barrier for selling high-value Filecoin-stored data, but the grant does not commit to generating measurable mainnet deal volume, FIL payment activity, FVM transactions, or paid flagship-client adoption by the end of the project. 3.3 References My latest collaborator from industry: Clara Shikhelman <[redacted]> Head of Protocol Research at [[alloc] init] My former graduate student: Alireza Arjmand <[redacted]> Head of Security Research at Cantina 4.1 Monthly Operating Burn [ "$10-$100K (small team)" ] 4.2 What % of total team monthly burn depends on this grant? 25-35% 4.3 If this grant is not awarded, what happens? If this grant is not awarded, the existing research and high-level public prototype will remain available, but the Filecoin-oriented implementation will likely not be completed on this timeline. We would not have dedicated support for the graduate student implementation effort, Filecoin/IPFS integration, benchmarking, documentation, and release engineering. 4.4 Core Team The core team consists of Prof. Majid Khabbazian as PI and technical supervisor, and one graduate student implementation lead who will work full-time on the project. Prof. Khabbazian is the single author of “Plaintext-Scale Fair Data Exchange,” accepted at FC 2026, and will provide protocol guidance, architecture review, milestone supervision, and technical validation. The implementation lead is a graduate student at the University of Alberta with industry software engineering experience. They will be responsible for the core implementation, seller/buyer CLI tools, Filecoin/IPFS-compatible demo, testing, benchmarks, documentation, and release engineering. Additional graduate students in Prof. Khabbazian’s group may provide limited support through code review, testing, benchmark feedback, and discussion, but the main delivery responsibility will rest with the PI and the dedicated implementation lead. 4.5 Has your team received a ProPGF grant or funding from PLFIF before? [ "No" ] 5.1 Key risks & dependencies Key risks are implementation complexity, scope control, and Filecoin/FVM integration uncertainty. We mitigate these by limiting the first grant to a core open-source implementation, seller/buyer CLI tools, a Filecoin/IPFS-compatible demo, and a modular settlement abstraction rather than a full production marketplace or mainnet payment system. The main external dependency is access to Filecoin/IPFS-compatible storage and ecosystem feedback. The project can still be delivered with a reproducible local/mock settlement backend if deeper Filecoin-native payment integration is deferred. Team risk is limited because one dedicated graduate student will lead implementation under direct PI supervision. Objective 1 Indirect Objective 2 Indirect Objective 3 Indirect