[Proposal] Live Balancer contract segments visualization and subscriptions

[Prabhaav_SimpleID]

#The proposal in one sentence:

Live Balancer contract segments visualization and subscriptions

##What? What will be done, what are the deliverables

• We will build 5 custom segments of Balancer contract data that community members can visualize and subscribe to for regular updates. The subscriptions can be used for monitoring tools, custom trading bots, arbitrage opportunities, dashboards, etc.

• The community can request and vote on the custom segments we will build (The data must be on-chain). Potential segment ideas that are of interest to include:

  1. Whale Watch: a breakdown of inflows and outflows of the top wallets holding Balancer tokens. Example use case - if a whale wallet always dumps their BAL on the market right after they claim, maybe traders want to be notified so that they can buy some discounted BAL.
  2. Monitoring smart pools parameters: a subscription continually monitoring smart pool parameters to keep community members, investors, and other interested parties aware of important changes. This requires parsing anonymous events.
  3. Dynamic smart pool fee adjustments: using live trading volume information, a smart pool controller could adjust the pool fees to maximize revenue (ie, if trading volume is high, increase fees; if low, decrease fees)
  4. Community Defined Segment - Pending Definition
  5. Community Defined Segment - Pending Definition

• We have already built out parts of the needed infrastructure and visualizations to support this proposal:

  • Webhook subscriptions for standard smart contract event monitoring
  
  • Continuous wallet balance updates
  
  • Active user information
  
  • Retention data
  

##Why? What problem it solves

• Easy access to structured high-performance blockchain data is imperative to promote new products being built on Balancer. Balancer segment data can help developers build trading/arbitrage bots, boutique financial instruments, and monitoring tools easily. We believe this will increase the liquidity available on Balancer while improving the efficiency of the protocol.

• We give developers an easy to use live segment data subscription service. This frees them from creating and maintaining custom sub-graphs or advanced-SQL to continuously monitor Balancer on-chain data.

• Multiple analytics/monitoring proposals in the community can take advantage of our service: https://forum.balancer.fi/t/proposal-marketing-growth-grant/201, https://forum.balancer.fi/t/proposal-dashboard-to-view-the-bal-community/248, https://forum.balancer.fi/t/liquidity-vision-balancer-grant-proposal/464

##When? Timeline and milestones with the requested payment schedule in BAL (subject to vesting) or DAI/USDC

Phase 1: Build Custom Segments & Multiple Subscription Service

• Est. Completion: March 2021
• Deliverables
  • https://balancer.simpleid.xyz
    • 5 custom segments
    • Segment visualizations updated every 20 minutes
    • Allows viewers to subscribe to updates for each segment
      • 20k Max.
      • First come, first served
      • 3-failed requests drops subscription
• Budget:
  • 2.5K Balancer Tokens vested over 12 months
  • Covers infrastructure costs for 12 months
  • Segment Pricing: $400 / month (5-segment pricing), Includes
    • 20 minute updates
    • 20 minute visualization updates
    • Server, Compute & DB Instances
    • Daily wallet & token balance updates
    • Data transfer
    • Storage

Phase 2: Live Updates (1-3 Block Delay)

• Est. Completion: June 2021
• Deliverables (beyond phase 1)
  • https://balancer.simpleid.xyz
    • 5 additional live update custom segments
    • Segment visualizations updated every 1-3 blocks
    • Allows viewers to subscribe to updates for each segment:
      • 20k authenticated URLs for high-performance updates (updated on avg. 15-45s)
      • 100k for regular perforamance updates (every 20 min.)
• Budget:
  • 12K Balancer Tokens vested over 12 months
  • Covers infrastructure costs for 12 months
  • Segment Pricing $2000 / month (5-live segment pricing), Includes:
    • 1-3 block updates
    • 1-3 block visualization updates
    • Server, Compute & DB Instances
    • Wallet & token balance updates every minute or faster
    • Performance replication costs
    • Data transfer
    • Storage

##Where?

The project will be available at https://balancer.simpleid.xyz

##Who? Team members and their roles in the project and backgrounds

Prabhaav Bhardwaj
Role: Software Engineer, UX/UI Designer
GitHub: https://github.com/prabhaav

Background:
Co-founder at SimpleID
10+ years of SW Development

Alex Carreira
Role: Systems Engineer
GitHub: https://github.com/AC-FTW

Background:
Co-founder at SimpleID
20+ years of HW(FPGA) & SW Development

[tongnk]

My thoughts on this is that I’m really unsure about the use case and especially the cost proposed. Given most things are built on top of subgraph - asking devs to build and maintain their own databases seems to be a bit counterintuitive vs utilising subgraph.

From an analytics project perspective we wouldn’t be able to utilise this as we rely on subgraph right now and don’t plan on storing the data ourselves (as it gets very expensive). I’d love to be proven wrong and we would be able to utilise this without moving from subgraph though.

Also for clarification - segment pricing $2k p/m is that Segment the software (segment.com) or is that something else

[Prabhaav_SimpleID]

Thanks for your perspective @tongnk, and I totally agree with you that the Graph could satisfy some use cases. However, what we’re offering are complete solutions that add value to new smart pool managers, LPs, traders, investors, community members, etc in a few ways:

• flexibility to add features that require external data and non-chain data (i.e. price feeds):

  • composability with other protocol data i.e external on-chain data not available in Balancer sub-graph as mentioned here: https://docs.balancer.finance/guides/smart-pool-templates-gui/perpetual-synthetic-pool
  • modify segments to fit community needs better within reason once a month: i.e. include comparisons to other AMMs already in our system or adding certain AMMs, not in our system at our discretion, modifications to event update format, etc.

• monitoring smart pools and alerting system for timely notifications:

  • alerts for pools being drained, see previous incident: https://medium.com/balancer-protocol/incident-with-non-standard-erc20-deflationary-tokens-95a0f6d46dea
  • alerts for parameters being changed as mentioned here: https://docs.balancer.finance/smart-contracts/smart-pools/rights-manager
  • off-chain monitoring and updates based on volatility for swing trading mentioned here: https://docs.balancer.finance/guides/smart-pool-templates-gui/swing-trading-pool
  • alerts for large balancer token trades when whale wallets are moving assets
  • webhook (push) vs rest API (pull): faster responses for alerts instead of building polling infrastructure

• pre-calculated metrics and visualizations:

  • engagement/retention in Balancer vs other AMMs to understand product/market or better yet smart-pool/LP fit
  • understanding investor behavior by visualizing asset inflows/outflows, stickiness, etc
  • arbitrageurs seeking profit through market inefficiencies between DEXs or CEXs
  • visualizing trends of major transactions/timelines, movements in whale wallets

• infrastructure benefits:

  • decentralizing critical infrastructure from a single point of failure (the Graph)
  • mitigating development costs and lowering the barrier to entry for asset managers
  • new smart pool LPs don’t have to build custom monitoring/visualizations or alert systems

As for the pricing you mentioned. The $2k/month (200 BAL/month/segment) was a breakdown of per segment pricing for the number of Balancer tokens (12k BAL) we requested for 12-months. It has nothing to do with segment.com.

What has your experience been developing with the Graph in terms of adding new features, expanding the subgraph, or joining data across different protocol subgraphs in a query?

[markus]

Thanks for the detailed proposal, @Prabhaav_SimpleID

Push notifications for state changes seem like a very useful feature at the base layer, but I hope we can hear from the Balancer community what they would like to build or see others build on top of this. This will help us more objectively scope the proposal (ie. clearly define the “segments” that will be built) so that everyone is on the same page from the start.

[Prabhaav_SimpleID]

Thanks, @markus. I agree we want to build the segments & visualizations that add the most value to the community. It looks like some of the work we’ve proposed is already of interest:

• #3 (dynamic smart pool adjustments) of the custom segments mentioned in my initial post is now a bounty offered: https://github.com/balancer-labs/balancer-bounties/issues/2.

• Advanced data analytics of Balancer usage, mentioned in @immutbl’s blog post: https://medium.com/balancer-protocol/batch-2-of-balancer-ecosystem-fund-grants-is-now-available-c6b1bf567483

[immutbl]

Hey @Prabhaav_SimpleID, appreciate your patience and apologies for the delay.

The feedback I’ve received on this proposal is that the visualizations proposed can be done on Dune Analytics, and the key differentiator in this tech seems to be the ability to subscribe to specific state changes within Balancer - eg weights/fees being updated, large swings in spot price. Ideally this proposal would be coupled with another one that makes use of the tech on that front, not just for visualizations.

So as this proposal stands, I’m going to decline this.

I highly appreciate your proposal and I would encourage you to consider Markus’ suggestions both here on the forum and in your 1-on-1 discussions if you’re interested in continuing to pursue a grant for this idea. Also would love to see proposals for any other interesting ideas you may have.

Thank you!

[Prabhaav_SimpleID]

Understood @immutbl. Will work with @markus on fleshing out the details for the smart pool monitoring system for a different proposal.