<aside>
💡 Notion Tip: Create a new page and select Daily Entry ****from the list of template options to automatically generate the format below every day.
</aside>
This was a very special week, hence I'll shift from my standard weekly update format. I don't think I can capture all that happened this week in the rigid form of a standup, so I'm "letting my imagination run wild".
Still reeling from the previous week's workload and the steep learning curve of DeFi protocols, we began the week with webinars by the folks from Chainlink. Chainlink's "Using any API" feature immediately jumped out to me as the magic lamp, and I couldn't help but wonder the range of opportunities this provides to a DApp developer. For someone with background in Backend web development, I was finally in my comfort zone, and I could let my imagination run wild. Call it coincidence or whatever, but while going through my repositories that night, I revisited this old project called VehiCO that I had built during the MLH Fellowship. VehiCO keeps track of the user's trips and their carbon emissions and lets the user offset by donating to NGOs, but we had left the donation feature of the app incomplete back then.
https://github.com/MLH-Fellowship/VehiCO
I kept wondering if it would be possible somehow to integrate a DeFi donation service with this app or build something similar that would be beneficial for a social cause.
There were 2 more webinars, where I learned about Biconomy and meta-transactions, and another about Layer2 and Matic, which only fueled my interest in DeFi and DApp development.
Equipped with the tools needed to build our first DApp, we were suggested to participate in the Etherpunk hackathon organized by Devfolio which had been running for a couple of weeks. It was suggested to us that we can submit the projects we had been working on since the previous week, but I wanted to try something new 😎
With only 4 days remaining to submit our projects to the hackathon, some of us decided to team up to split the work so that we could build something substantial. I teamed up with Ankur from my cohort(Team Divya) and started brainstorming ideas. The next 2 days involved a lot of Eureka moments and weighing a lot of ideas, and we struggled to come up with a feasible solution where we could use the protocols we had learned about in daily life. We scrapped my idea of extending the VehiCO project as it would involve keeping track of how much each user contributes to harming the environment, and we weren't sure where to get that data from. However, this gave us another idea: VehiCO had involved a lot of data fetching from Weather and Location APIs, and Chainlink oracles could read data off any API. Around Friday morning, after around 2-3 calls and a lot of research on how health and life insurances work, the concept of HurriCARE was born - a decentralized insurance platform where users would pay a premium based on location and would get instant payout in case a cyclone or a hurricane actually struck that location.
We quickly decided on the protocols and tools we'd need for the project: Chainlink for oracles, Chainlink Alarm Clock for scheduled periodic checks, Superfluid for streaming premium payments, Matic for L2 deployment and Portis/Metamask for login. I decided to take up researching Chainlink, and Ankur researched Superfluid, and we prayed and hoped we could get the two contracts for premium payment and checking for payout ready by Saturday so that we could devote Sunday to integrating Matic and Portis.
By Saturday night (I should say Sunday morning coz none of us slept 😅), Ankur had a barebone React UI ready with MetaMask login and Superfluid stream, and I built an API to accumulate the data in the required format from OpenWeather API that the Chainlink oracle would read off from.
Things didn't go exactly per plan (in my experience they never do in hackathons), and it was only around 5pm on Sunday that we were confident our standalone contracts were working on testnets. Then came the challenge of integrating the two contracts together, deploying them on Matic and get Portis (and possibly Biconomy) working. Each of them presented its own challenge as we were just beginners in writing Solidity code, and to our utter dismay we found that some of the tools we were using were not available across networks - Chainlink Alarm Clock oracle was not deployed on Matic Mumbai Testnet. So we commented it out and for the demo, we ran a custom web3 script to trigger the payout. By 8pm, we had deployed on Matic and were debugging Portis login. Then disaster struck again - as I was building the README for the project repository, Ankur's power went out. You'd ask how that counts as a "disaster" given that we were 2 people and any of us could record the demo video, but here's the thing: my mic was broken, and whenever I unmuted myself it'd sound as if I were in a warzone. We came up with a backup plan that I'd screen-record the working application, and Ankur would lend the voiceover, but thankfully his power came back soon, and he was able to record the demo video. With hardly 15 min to go, we submitted our project for the Matic, Chainlink, Superfluid and Portis bounties and heaved the first sigh of relief in 3 days.
Needless to say, we had little to no expectation out of our submission, especially after we saw the level of cool projects that were submitted. The following Thursday, we got reached out to by the SuperFluid community and they wanted us to present their idea at the Superfluid Community Call. Our project was largely appreciated by the Superfluid community and we got some valuable feedback regarding future prospects.
People say all's well that ends well. HurriCARE won the Runners-Up prizes for both Chainlink and SuperFluid bounties (a whopping $500 in prize money for each). But regardless of the prize, what mattered to us was the recognition our idea received from the experts, and the fuel it provided to our passion to keep BUIDLing.