Blockchain data mining: When adopting blockchain solution into actual use case, the characteristics of data change adding real-time, global and authorized/notarized features. Explore the business value of Blockchain data by use case.
- E.g. Worked for FX trade transaction across internal offices -> built on R3 Corda -> payment (SWIFT msgs) data that was stored fragmentedly in each office becomes real time, global and settled (notarized by Central bank) data, Blockchain data.
- Based on the Blockchain data, real-time liquidity flow can be predicted (what amount of liquidity is expected at which currency & which corridor) or real-time risk position calculation is possible by office and global position
Blockchain data science stack & tool development: new tools for blockchain data science stack, ETL, Computation, Visualization should be re-designed & developed to best perform with Blockchain tech stack/system
- ETL: how to create global datasets given the limitation of data protection acts and interest conflict among banks
- Computation: on-chain/off-chain
- Visualization: generalized tool built on the EEA solution
Three models are possible for creating global data for analytics.
- SGX + ZKP type query system
2. Virtual node which aggregate global data
3. Data governance model (no drawing yet)
Data governance architecture design & development: Tackling ETL issue first
- Goal:Create real-time global(shared) data system for analytics
- Two (simplified) challenges: 1) Data protection and residency laws compliance and 2)
- Questions: How much data should be shared, how to be shared and where to be stored
- The idea started from a chat with Tendermint team last week on how to utilize Cosmos hub for banking/enterprise solution. As far as I understand from the chat with them (just started reading its WP), their new product, Cosmos Hub, is an interoperability solution. In the Cosmos hub, each zone can be interoperable through hub and cosmos hub is strong at flexibly configuring governance of the zone. (might need a new technology to actually implement)
- Context: I am currently working on a project where I need to transfer some customer data from HK office to UK office, there are massive compliance works required and 6 lawyers & compaliance experts started looking into it… for one data transfer
- My idea is building data governance structure for each zone (possibly each country) and build gateway between the zone to transform the data (real-time notarized by regulator in terms of data compliance or by a smart contract certified)
- Goal is building maximum amount shared data between organizations (zones) without data related compliance risk
- Also each corridor or each region (where the compliance is similar) can share more data than global data. That is, it’s for data sharing across zones to comply to regulatory rules, and shared data pool for each corridor can be optimized (sharing the maximum amount of data just marginally meeting the compliance)
- Five methods of sharing data across organizations
- Depending on corridors, certain data fields can be shared in one of methods above, the method choice can be made
- This can be built with enterprise dlt solution together as data governance solution
- Two use cases can be suggested:
- Intra-transactions for internal office which have different data protection compliances
- Inter-organizations (competition suppress data sharing)
- Next step1: Idea validation (business)
- What’s the specific use case for data sharing for internal offices and inter-organizations?
- What’s the current data sharing/transfer process for banks? Why the current data sharing/transfer process is complicated/expensive?
- Expected benefit to address this issue by use case
- Next step2: Idea validation (Technology feasibility)
My (technical) interest on DLT is analytics on DLT, and first challenge I found is securing/creating the global datasets to analyze.