I will build an exporter on yearn to serve as an advanced data visualisation tool for yearn.vision. The first task will consist of building a statistics and machine learning dashboard in Grafana.
The second task will involve all current dashboard sections to be enhanced and updated, [ General, Strategists, System, Treasury, Treasury-dashboards, Yearn] to include customized graphs with statistics and metrics each team in yearn can utilize built to their specific needs.
Plan
I have coordinated with the exporter team to identify a practical use case for the dashboard that all teams can benefit from. I will be constantly working with the exporter and other teams to ensure all the applicable steps are taken to conduct this project in a sequential manor.
The 3-month plan is as follows:
PART 1
1.1 I have spoke with members of the exporter team and they would benefit from the following:
a). The exporter team pull GBs of data from the chain + from the vaults but only a small amount get analysed, " i would rectify this by running data optimization techniques using ML to filter data that can be used as an advantage.
b). Filter and improve the overall data availability the exporter team use and other may use.
1.2. From yETH team they would benefit from the following:
a). Yield over time and the contribution of the different sources (rewards, tx tips, mev)
b). Validator performance, i.e. slashing rate, missed attestations, average attestation inclusion distance
c). OFAC%
d). Consensus/execution software distribution for each (if available),
1.3. From yETH team they would benefit from the following:
a) stETH vs ETH volatility history in different statistical methods
b) GMX's (on Arbitrum and maybe Avalanche too) GLP make-up distribution between BTC, ETH etc. in historical context
PART 2. Design and create a Grafana statistics dashboard page under yearn vision with the following statistics + feedback from teams.
• Teams preferred statistics / metrics.
• Correlations with vaults, share price and ETH, BTC, etc.
• Multiple types of Regression analysis with vaults, share price and ETH, BTC, etc.
• Heatmap of gas prices.
• Extension of share price descriptive statistics
• Update to version 10 for Grafana time series charts.
• Machine learning forecast.
• Update Prometheus with new metrics.
(example image of grafana ML prediction Dashboard)
3. Enhancing the current dashboard sections of yearn vision will consist of the following:
• Speak to each team to identify what data visualization statistics or metrics they can use to their advantage.
• Coordinate across teams to update each sections dashboard template layout.
• Implement new charts based of feedback.
4. Write a report of findings (end of each month)
Yearn teams will benefit by:
• Metrics will be built up based on research and feedback from teams, to ensure they maximise utility and causality from data.
• Test variables on a quantitative and qualitative level to gain a deeper understanding to what extent variables influence each other for instance. E.g., a team member asks for a custom statistic to be made with deposits and APY to forecast profitability.
• By having an advanced data tool to visualize on-chain statistics and metrics, yearn would be able to break down variables such as total TVL, and understand how correlated it is with individual assets in the DeFi space subject to seasonality (Bull + Bear runs).
• Alerts will be implemented from the machine learning forecast variables subjected to teams’ preference, for instance alerts can be programmed subject to large moves in a forecasted variable.
Value to yearn overall:
• Build custom metrics like forecasting will contribute towards future profits.
• Lay the foundation to be the home for all yearns future statistics e.g., V3, yETH metrics.
• Yearn to be one of the first DeFi protocols to test and obtain machine learning algorithms (with on-chain data) which in turn will contribute value by optimizing decision making.
Deadline
2023-14-07
People
CSdata - Responsible for creating and updating dashboards.
additional yearn contributors not covered in this budget request.
Money
I am requesting $2500 DAI a month for 3 months totalling $7500 DAI. The money will be used to conduct all tasks described in the plan.
Scope
I will build an exporter on yearn to serve as an advanced data visualisation tool for yearn.vision. The first task will consist of building a statistics and machine learning dashboard in Grafana.
The second task will involve all current dashboard sections to be enhanced and updated, [ General, Strategists, System, Treasury, Treasury-dashboards, Yearn] to include customized graphs with statistics and metrics each team in yearn can utilize built to their specific needs.
Plan
I have coordinated with the exporter team to identify a practical use case for the dashboard that all teams can benefit from. I will be constantly working with the exporter and other teams to ensure all the applicable steps are taken to conduct this project in a sequential manor.
The 3-month plan is as follows:
PART 1
1.1 I have spoke with members of the exporter team and they would benefit from the following:
a). The exporter team pull GBs of data from the chain + from the vaults but only a small amount get analysed, " i would rectify this by running data optimization techniques using ML to filter data that can be used as an advantage. b). Filter and improve the overall data availability the exporter team use and other may use.
1.2. From yETH team they would benefit from the following:
a). Yield over time and the contribution of the different sources (rewards, tx tips, mev) b). Validator performance, i.e. slashing rate, missed attestations, average attestation inclusion distance c). OFAC% d). Consensus/execution software distribution for each (if available),
1.3. From yETH team they would benefit from the following:
a) stETH vs ETH volatility history in different statistical methods b) GMX's (on Arbitrum and maybe Avalanche too) GLP make-up distribution between BTC, ETH etc. in historical context
PART 2. Design and create a Grafana statistics dashboard page under yearn vision with the following statistics + feedback from teams.
• Teams preferred statistics / metrics. • Correlations with vaults, share price and ETH, BTC, etc. • Multiple types of Regression analysis with vaults, share price and ETH, BTC, etc. • Heatmap of gas prices. • Extension of share price descriptive statistics • Update to version 10 for Grafana time series charts. • Machine learning forecast. • Update Prometheus with new metrics.
(example image of grafana ML prediction Dashboard)
3. Enhancing the current dashboard sections of yearn vision will consist of the following:
• Speak to each team to identify what data visualization statistics or metrics they can use to their advantage. • Coordinate across teams to update each sections dashboard template layout. • Implement new charts based of feedback.
4. Write a report of findings (end of each month)
Yearn teams will benefit by:
• Metrics will be built up based on research and feedback from teams, to ensure they maximise utility and causality from data.
• Test variables on a quantitative and qualitative level to gain a deeper understanding to what extent variables influence each other for instance. E.g., a team member asks for a custom statistic to be made with deposits and APY to forecast profitability.
• By having an advanced data tool to visualize on-chain statistics and metrics, yearn would be able to break down variables such as total TVL, and understand how correlated it is with individual assets in the DeFi space subject to seasonality (Bull + Bear runs).
• Alerts will be implemented from the machine learning forecast variables subjected to teams’ preference, for instance alerts can be programmed subject to large moves in a forecasted variable.
Value to yearn overall:
• Build custom metrics like forecasting will contribute towards future profits. • Lay the foundation to be the home for all yearns future statistics e.g., V3, yETH metrics. • Yearn to be one of the first DeFi protocols to test and obtain machine learning algorithms (with on-chain data) which in turn will contribute value by optimizing decision making.
Deadline
2023-14-07
People
CSdata - Responsible for creating and updating dashboards. additional yearn contributors not covered in this budget request.
Money
I am requesting $2500 DAI a month for 3 months totalling $7500 DAI. The money will be used to conduct all tasks described in the plan.
Amount
7500 DAI
Wallet address
0x6e7232AD524a23CF623410D4b214a434A90d61bB
Reporting
Weekly