Inspiration

Since launching, prediction markets have been viewed as an alternate/unregulated avenue for gambling. People underestimate their "predictive" accuracy. We built this to highlight what we believe is a compelling use-case for prediction markets: risk management.

What it does

The user (small business owner, fund manager, investor) enters their financial risk / position in plain English. Vectis finds the relevant prediction market(s) and proposes position(s) that could offset some of that exposure. For example, a business with a floating-rate loan could enter a position that pays out if the Fed holds rates steady. You can review the reasoning, edit the size and leverage, or remove any position before signing. This is built on top of our broader platform that allows leveraged positions on prediction markets for more intense hedging or stronger market theses.

How we built it

We built an agent using Venice (for private inference) that searches Gemini’s market catalog and reads its order books. Our backend checks the proposed positions against Gemini data and applies the contract’s leverage limits. On Arc, an ERC-4626 vault supplies USDC, while a margin contract handles collateral, liquidation, and settlement. QuickNode Streams delivers filtered contract events to our activity feed and the keeper that processes positions.

Challenges we ran into

Nailing the UX for this project was a big challenge since there are so many critical parts coming together: the user's natural language risk description, surfacing the relevant markets, and letting the user customize their sizing and positions. Combining these in a sleek, intuitive interface was hard but after some iteration, we arrived at the final flow which intuitively shows the positions and sizes and allows for edits before confirming.

Accomplishments that we're proud of

As mentioned above, achieving a smooth UX is something we're proud of. Also integrating the natural language aspect to it was nice as it just further simplifies the product. Vectis solves a genuinely painful problem in a way that is super easy to use. We also implemented the leveraged position lifecycle on Arc testnet, from borrowing USDC to repaying the vault and returning the remaining proceeds to the trader.

What we learned

Through this project, we were able to explore the actual implementation mechanics of leveraged positions onchain and how smart contracts interact to begin, monitor, and liquidate these leveraged positions. We were also able to get more familiar with various types of hedging, different ways financial risks can arise, and how hedging is traditionally done today.

What's next for Vectis

I see several paths ahead for Vectis. First, I would want to implement more financial instruments. Onchain stocks, options, derivatives, etc. could make for more diversified hedging basket. Furthermore, implementing broader portfolio management and asset tracking to directly track risks (rather than having the user identify and define them). At that point, it would become an intelligent manager to identify financial risks itself, suggest investments and hedges, and divert capital towards various onchain asset types.

Built With

Share this project:

Updates

Submission history