anyway.

Even experienced operators have a hard time to keep going on high demand peaks: check the explanation [1] by Sebastien Guillemot from Emurgo for the incident Yoroi backing services recently faced. It was caused by simultaneous NFT projects launches that highly increased transactions traffic in the network.

To build a more decentralized solution, we basically face 2 problems:

• It's actually a challenging task to setup every available Cardano API and deal with its updates
• In addition, there is a lack of incentives for operators to deal with these challenges.

There is also a need to build this network along with a reputation system based on availability and performance metrics that can be used and trusted by API consumers to choose the most suitable deployment for them.

References

[0] "If we don't stop relying on Infura, the vision of ethereum failed" https://www.coindesk.com/the-race-is-on-to-replace-ethereums-most-centralized-layer [1]

DETAILED PROPOSED SOLUTION

Desired Outcome

The desired outcome for this proposal is to take to the next level the original "Cardano API as a Community Service", which was merged into Gimbalabs and voted as winner on fund2. Dandelion free community service has facilitated bootstrapping more than ten projects since the beginning of Fund2 such like:

• SpaceBudz [0]
• AdaBoost Cardano Calendar [1]
• lift-wallet [2]
• raywallet [3]
• MONAD's pool network observability tool [4]
• Blocktales interactive book [5]
• Cryptomayor NFT portal [6]
• LacedWhales Twitter bot [7]
• Cardano Metadata Viewer by Icaro Capobianco [8]
• Development courses where it was used as learning resource (ADOPT [9] and PoA [10] dev courses)
• Maybe others use it silently, we preserve privacy :)

Now these and many others are in stages where they might be in the need of backing services that support their production workload, it would be nice to start building this market offer in a standardized way so projects could easily switch between deployments or just spread their load.

The plan is to nurture a network by developing the tools needed to ease the deployment and maintenance and to incentivize SPOs (the most natural builders for it) or any other operator by having an easy way to monetize such services using Cardano native assets.

Solution overview We'll focus on building further the technology required to make Dandelion a self-servicing API where any consumer can buy and employ Cardano native assets as master key to access the backing service. The piece of software that makes this possible has been proposed independently in Fund5 metadata challenge [11] as a generic blockchain authorization mechanism and we'll be integrating it into Dandelion for the purpose described below.

On the other hand, to enable API consumers to choose the most suitable deployment, such distributed network of API deployments will require a trustable reputation system, so we'll start an in-depth research to build this solution. In addition, we'll be launching a campaign within the SPOs community to bootstrap the network/market with as many deployments as possible.

We, of course, will keep offering support from Gimbalabs' discord server [12] about the usage, deployment and maintenance of these services and improving the Infrastructure as Code templates (kustomize-dandelion [13]), updating them to support latest service versions and including new ones like blockfrost's [14], plutus-pab or ipfs-cluster (we've already started integrating [15]) that help projects adopt the decentralized revolution on top of Cardano in a friendly manner.

3 months roadmap tentative

• Setup staging environments and maintain 2 different production versions to help projects handle migrations
• Deploy on our behalf in at least 1-2 new locations to provide a minimal level of redundancy to the network
• Look for SPO candidates to deploy on extra locations
• Start promoting the initiative inside and outside Cardano community by writing proper documentation and storytelling
• Integrate the Fund5 proposal work on an authorization mechanism based on Cardano native assets
• Start researching and planning an MVP for the reputation system

Costs estimates We identify that to push the initiative to the next level, we'll need at least 4 dedicated roles:

• 1 Technical Project Lead position that takes ownership of the product development and leads integrations ($2000/mo x3 months: $6000)
• 1 Systems administrator/DevOps advocate that takes ownership of the community service deployments and helps developing the opensource projects ($2000/mo x3 months: $6000)
• 1 Community Manager position that starts building the brand in social media ($1000/mo x3 months: $3000)
• 1 Technical "CEO" (the proposer) that helps building/leading and looks after for business sustainibility ($1000/mo x3 months: $3000)

Then we also identify these other budgets:

• Budget for operational costs for at least 2-3 locations and different versions for the community service, including staging environments (3x deploys $500/mo x3 months: $4500)
• Budget to help some experienced SPOs to bootstrap their own deploy (3x deploys $300/mo x3 months: $1800)

Total estimate: $24300 for a 3 months period.

Intellectual Property Every asset developed will be released under Apache 2.0 license

References [0] https://spacebudz.io/ [1] https://dbooster.io/calendar [2] https://github.com/CodingOnChain/lift-wallet [3] https://rraayy.com/ [4] https://monadpool.com/cardano.html [5] http://testnet.blocktales.dehlic.it/ [6] https://cryptomayor.io/ [7] https://twitter.com/LacedWhales [8] https://icaro-capobianco.github.io/cardano-metadata-viewer/ [9] https://discord.gg/FkB82BVYJH [10] https://discord.gg/eXCabdCPqm [11] https://cardano.ideascale.com/a/dtd/Blockchain-Authentication-Mechanism/352005-48088 [12] https://discord.gg/qDc3f9R7Ab [13] https://github.com/GimbaLabs/kustomize-dandelion [14] https://blockfrost.io/#FAQ [15] https://github.com/GimbaLabs/kustomize-ipfs-cluster