search

livepeer / grants

⚠️ DEPRECATED ⚠️ Please visit the new homepage at https://livepeer.notion.site/Livepeer-Grants-Program-f91071b5030d4c31ad4dd08e7c026526
44 stars 7 forks source link

VideOracle - Grant Application #70

Closed codethazine closed 1 year ago

codethazine commented 2 years ago

Give a 3 sentence description about this proposal. VideOracle is a verification protocol that lets users earn rewards for helping to improve products, services, or their community. At the Berlin hackathon with Livepeer, we demonstrated how we can use video to verify real-world events in a smart contract.

Now we are seeking a research and development grant to further develop this technology to build a working, deployable video oracle protocol on Livepeer that developers can use to verify real-world events.

Describe the problem you are solving. Oracles provide data to blockchains to verify real-world events, and the type and quality of data provided by the oracle are essential. Video could provide a novel input for oracles that could open countless new use cases, and this is why we’ve developed VideOracle.

Describe the solution you are proposing. A reliable oracle requires decentralized sources of data to verify real-world events. Livepeer gives us a performing and decentralized layer on which to build a video oracle.

We are seeking to:

  1. Explore use cases when video might be the best source of data for an Oracle
  2. Create a protocol that other developers can use to verify real-world events with on-chain video
  3. Grow a user community to test and implement video oracle use cases

Describe the scope of the project including a rough timeline and milestones

Alpha app features:

The goal of the alpha app is

This project is split into three phases and tied to a successful outcomes that would benefit the Livepeer ecosystem. Phase 1 is split into three milestones and is focused on market research and MVP development. Phases 2 and 3 are focused on feedback iteration and user base growth.

Phase 1 (Project Launch) - $15k July - September 2022 Use case validation, MVP and demo on Livepeer community call

Milestone #1 (Estimated Deadline - July 2022)

Milestone #2 (Estimated Deadline - August 2022)

Milestone #3 (Estimated Deadline - September 2022)

Please estimate hours spent on the project based on the above:

  1. Market research: 80 hours total
  2. User on-boarding: 40 hours total
  3. Smart contract oracle system: 80 hours total
  4. Livepeer network integration: 40 hours total
  5. React Native app + UX/UI designs: 240 hours total

Phase 2 - $10k September - October 2022

Phase 3 - $10k October - December 2022

Amount of LPT Grant we are looking VideOracle is looking to obtain a grant amount of $35,000 separated into 3 phases ($15,000 on kickoff and two $10,000 contingent on milestone objectives).

Side note We started working on VideOracle at the NFT Berlin Hackaton, where we were selected by Livepeer as one of the winning projects. These are the deliverables we submitted:

  1. Presentation: https://www.youtube.com/watch?v=WXSrQuJYHvI&feature=emb_title
  2. Slidedeck: https://docs.google.com/presentation/d/1xNyiara9hMTQlsokC9InK_YD-Y2n-AQbj0xFgm2A_Ak/edit?usp=sharing
  3. Github: https://github.com/emilianobonassi/videoracle-dapp
nelsorya commented 2 years ago

Hey @codethazine, thanks for putting this together, this sounds like an interesting idea. Could you tell us more about how this would be technically implemented or provide a spec for this?

codethazine commented 2 years ago

Hey @nelsorya of course! :)

So basically this is the architecture we've implemented at NFT Berlin:

Schermata 2022-06-29 alle 10 48 57

So we have two parties:

  1. Requesters: Who make their requests through our DApp
  2. Verifiers: Who verify the requests through video

After a certain time window has passed, requesters distribute 5 votes to all the verification videos, and that triggers the smart contract to distribute the bounty among them. This is the Smart Contract we've implemented: Smart Contract on Mumbai net

All of this has been done on a WebApp in Vue, but we want to refactor the app to React Native, in order to give the possibility to verifiers of verifying videos directly from their smartphone.

The DApp, even though not fully functional, is currently running on Vercel: VideOracle But we want to migrate our servers to AWS Amplify for our backend, in order to have a more scalable and robust infrastructure.

Let me know if you have any other questions.

nelsorya commented 2 years ago

@codethazine thanks for sharing this additional context. There are some interesting concepts here and we would be interested in funding a research grant first for $5k in LPT first to further explore how the design for how this could work and gather feedback before turning it into a product.

We were thinking something of the style of https://ethresear.ch/t/proof-of-efficiency-a-new-consensus-mechanism-for-zk-rollups/11988 or https://ethresear.ch/t/mev-resistant-zk-rollups-with-practical-vde-pvde/12677

codethazine commented 2 years ago

Hey @nelsorya, sounds great! We could see this as a phase 0 to explore the technical implementation for a full-blown L2 video verification protocol.

Our current implementation puts the requester as the single source of truth, leaving space for exploitation. Adding 3rd party voters could give the architecture a reliable decentralized verification system for the videos. We'll need to allign the incentives for all the parties and see what consensus mechanism is ideal for our use case.

This will be our timeline for the next 6 weeks: W1 - Architecture Design W2 - Architecture Design W3 - Development W4 - Development W5 - Testing W6 - Deployment

Btw we would love to hop on a call in the following days and discuss what you had in mind and expect for this phase 0.

nelsorya commented 2 years ago

@codethazine sorry for the delay in following up, I missed this response. Ok great I'm Nelson#6080 on discord. Let's chat there and we can create a group and discuss next steps.