Ajey
Agentic DeFi Yields, Now on Octant!
About
The problem it solves # Problems Solved & Use Cases ## The Problem * **Public goods funding gap** Public goods struggle to secure *recurring* funding. Traditional donations are one‑off; DeFi yield is typically kept by users or protocols. Funding is often **single‑cause**, missing the chance to support a *portfolio* of public goods. * **Friction in impact investing** Many people want to support causes **without sacrificing principal**. Existing flows force a trade‑off (yield *or* impact), involve manual harvesting + sending, and introduce cognitive overhead that makes recurring impact unlikely. * **Fragmented DeFi infrastructure** Building yield strategies across assets requires swaps, market selection, and protocol integration. Cross‑asset allocation typically implies **custom code, higher surface area, and audit burden**. * **Trust & transparency** Donors want to know *where* funds go and confirm *that* donations occur. Multi‑cause distribution and the timing of donations need to be **provable on‑chain**. --- ## What Problems It Solves **Problem 1: Sustainable public‑goods funding** **Solution: Automatic, multi‑recipient yield donation** * **100% of realized yield** is donated on‑chain using preset **split mixes**: *Crypto‑Maxi* (60% crypto public goods / 20% humanitarian / 20% hygiene), *Balanced* (40% / 30% / 30%), *Humanitarian‑Maxi* (20% / 40% / 40%). * Donations recur as TVL accrues yield; no manual steps. Realization occurs at controlled checkpoints (harvest/fee‑take), then routed. **Problem 2: Capital preservation vs. impact** **Solution: Keep principal; donate only yield** * Users deposit into **ERC‑4626** single‑asset vaults (WETH/USDC/USDT/DAI) that supply to **Aave v3**. * Depositors keep their **original principal** (subject to protocol/market risks); **all net yield** is directed to public goods according to the chosen mix. * No performance/management fees to the team; the design focuses on **donation‑first economics**. **Problem 3: Complex DeFi integration** **Solution: Composable architecture + agent orchestration** * Instead of a heavy multi‑strategy vault, Ajey uses **four small, single‑asset ERC‑4626 vaults** + an **agent orchestrator** that: * monitors Aave v3 yields across **WETH/USDC/USDT/DAI**, * executes **migrations** between vaults (e.g., DAI → USDT) via a **whitelisted swap aggregator**, * enforces **slippage/deadline guards** and role‑gated execution. * This keeps code surface **minimal and auditable** while achieving multi‑asset optimization. **Problem 4: Trust & transparency** **Solution: On‑chain verification, explicit addresses** * All donations are **on‑chain to published recipients**; transactions and amounts are independently verifiable. * ERC‑4626 share/accounting and event logs provide a clear audit trail (deposits, harvests, migrations, donations). * Role‑gated actions and explicit allowlists (swap aggregator) reduce discretionary risk. --- ## Who Can Use It * **Impact investors** Preserve principal while delivering **automated, recurring, verifiable** donations to multiple causes. * **DAOs & organizations** Align treasury management with mission: deploy idle capital to Aave v3 and **stream yield** to public goods—without new operational overhead. * **DeFi users** Prefer a **hands‑off** experience: deposit once; the system reallocates between Aave markets and donates yield by default. * **Developers & builders** Compose on a standard **ERC‑4626** interface; integrate Ajey as a building block for impact‑driven apps and dashboards. --- ## How It Makes Existing Tasks Easier/Safer **Easier: Automated donation process** * *Before:* Manual yield harvesting; manual transfers; high chance of forgetting; per‑donation gas. * *After:* **Automatic donation** at each realization event; **no manual steps**; batched/efficient flows; predictable cadence. **Safer: Capital preservation** * *Before:* Donation‑focused products sometimes commingle principal and donations; unclear rules and costs. * *After:* Principal remains in ERC‑4626 vaults supplying **Aave v3**; **only yield** is routed to causes; **no team take** or hidden performance fees. **Easier: Integration & maintenance** * *Before:* Build cross‑asset logic from scratch; maintain complex debt/queue code; multiple audits. * *After:* **Single‑asset vaults** + **orchestrator** pattern; small contracts; standard interfaces; easier testing and review. **Safer: Smart‑contract execution** * *Before:* Ad‑hoc donation logic, manual mistakes, centralized gatekeepers. * *After:* **Role‑gated** orchestrator, **whitelisted** swap aggregation, **slippage/deadline** checks, on‑chain logs. **Easier: Multi‑asset composability** * *Before:* Integrate each protocol/asset pair separately; custom swap plumbing; brittle migrations. * *After:* **One ERC‑4626 pattern** across assets; orchestrator performs **asset swaps + reallocation** between Aave markets.
How it's made
Challenges we ran into ### Challenges I ran into * **Wrong contract addresses on a mainnet fork:** Deploys produced incorrect addresses due to a pre‑deploy **chain simulation**. Fixed by deploying with `--skip-simulation`. * **Multi‑asset vault complexity:** A monolithic multi‑asset vault added heavy debt/accounting overhead. I **pivoted to one YDS strategy with four single‑asset vaults** (WETH, USDC, USDT, DAI), preserving multi‑asset behavior via the orchestrator.