Loading...
Supporting developers, startups, and full-fledged companies building on Stellar and Soroban
© 2025 Stellar Development Foundation
Subscribe to the Newsletter
By Turbolong
Turbolong is a leveraged-yield application on Stellar that amplifies returns on Etherfuse stablebonds (CETES, USTRY, TESOURO), USDC, and XLM by executing recursive supply/borrow loops on Blend v2. Every loop, whether 3× on a Mexican Treasury-backed stablebond or 12.9× on USDC, settles atomically within a single pool.submit() call, so the user signs once and either the full loop lands or it reverts. This ergonomics is Stellar-native: on EVM chains the same outcome requires flash loans.
Turbolong is live on Stellar mainnet against three Blend pools: Blend-Etherfuse, Blend-Fixed, and Blend-YieldBlox (currently frozen following the February exploit). The application already composes Blend v2, DeFindex (vault strategy wrapper on testnet only), Soroswap (BLND harvest), and Stellar Wallets Kit (five-wallet sign-in surface) in production. Supporting infrastructure includes a Cloudflare Workers + D1 alerts service that monitors APY across pools and leverage brackets.
This application focuses narrowly on four Integration-List building blocks that, together, take Turbolong from its current live state (three Blend pools on mainnet, four Integration-List building blocks composed in production) to a composable DeFi primitive other Stellar applications can build on: DeFindex (deepened), Stellar Wallets Kit (deepened), Stellar Broker (new), and Aquarius (new). Blend v2 remains the core live dependency and is not itself an award deliverable. Soroswap stays in place as a fallback swap path once Stellar Broker becomes the primary router.
The DeFindex and Stellar Wallets Kit integrations carry the "deepened" tag because Turbolong already ships both in production today, built without SCF funding. The "deepened" label is not a request to fund existing work. The application’s DeFindex line items (T1–T2) fund five concrete net-new additions that do not exist on mainnet today: mainnet vault deployment across four assets, SEP-41 transferable receipt token, migration helper, partial-unwind liquidation protection, and auto-rebalance keeper. The Wallets Kit line item (T1) funds three concrete net-new additions: mobile deep-linking, Ledger hardware support, and a full E2E test matrix across all five wallets. Stellar Broker (T2) and Aquarius (T3) are net-new integrations from zero.
Every line in the deliverable maps to engineering work that does not exist on mainnet today; acceptance criteria are written against artefacts (deployed contracts, merged code, on-chain transactions, test reports) that will be produced during the tranche windows and verifiable on Stellar Expert and the public GitHub repository.
1) DeFindex vault strategy
Already shipped on Mainnet (no funding): non-custodial frontend over Blend that computes the leverage strategy (loops, HF, net APY), builds transactions, and lets the user sign them. b-tokens / d-tokens stay registered to the user's address in the Blend pool; Turbolong never holds funds.
Already shipped on testnet (no funding): Rust/Soroban strategy contract wrapping Blend loops into an ERC-4626-style vault via the DeFindexStrategyTrait. Users deposit once, the vault manages the loop, harvests BLND, and can rebalance. It's basically an automation of what currently exists on mainnet.
This award funds (net-new work): mainnet vault deployment across all four supported assets (USDC, CETES, USTRY, XLM — the address slot is currently a placeholder in frontend/src/defindex.ts), SEP-41 transferable receipt token (does not exist today), migration helper (does not exist today), partial-unwind liquidation protection (does not exist today; HF breach forces full close in the current build), and auto-rebalance keeper (does not exist today; rebalance is a CLI script).
2) Stellar Wallets Kit
Already shipped on mainnet (no funding): single sign-in surface across five Stellar wallets (Freighter, xBull, Albedo, Lobstr, Hana) covering classic and Soroban ops. All signing flows are kit-native.
This award funds (net-new work): mobile deep-linking for Lobstr / xBull / Hana so users can sign from phone wallets without copy-paste (does not exist today; mobile signing requires manual XDR copy-paste in the current build); Ledger hardware-wallet support via the kit adapters (does not exist today); a passkey-based smart-account path for returning users (stretch, scope-gated by kit support); and automated E2E test coverage across all five wallets and seven flows (does not exist today; coverage is partial).
3) Stellar Broker
Not yet integrated (zero baseline). Harvest swaps go through Soroswap only, which occasionally leaves value on the table when multi-source routing would beat a single DEX.
This award funds (net-new integration from zero): Stellar Broker integrated as the primary swap router for BLND harvest and any intra-vault swaps, with Soroswap retained as a fallback. The best-route choice is made per-tx from quoted output.
4) Aquarius
Not yet integrated (zero baseline). Turbolong currently reports only Blend-internal rates. Users have no way to compare against the wider Stellar DEX surface, and vault-share holders have no secondary-market exit.
This award funds (net-new integration from zero): (a) Aquarius liquidity data powering a 'Best Rate' badge in the Compare Pools view and 24h/7d rate-trend arrows, and (b) listing the SEP-41 DeFindex receipt token on Aquarius so depositors can exit positions without going through the full unwind path.
5) Blend v2 (core dependency — not an award deliverable)
Already live on mainnet (no funding): up to 20× theoretical / 12.9× practical leverage via atomic pool.submit() composing alternating supply and borrow ops. Uses Blend's b-token and d-token receipts for position tracking.
Not funded by this award. Blend v2 is a core live dependency, not an award deliverable. It benefits indirectly from better routing (Stellar Broker), better liquidity data (Aquarius), and a hardened vault (DeFindex).
6) Cloudflare Workers + D1 alerts (supporting infra)
Already live (no funding): 15-minute cron computes APY per pool/asset/leverage bracket and emails verified subscribers. Runs on the global Cloudflare edge with zero custodial surface.
This award funds (net-new additions to existing service): health-factor alerts (do not exist today), liquidation-imminent alerts at HF < 1.05 (do not exist today), and a historical rate-snapshot table (does not exist today) enabling the 7d/30d/1y charts that feed the Aquarius comparison view.
$99.0K
Team credibility: The Aha Company has a strong delivery record, including major contributions to the Stellar CLI, and institutional work such as bringing Société Générale-Forge’s EURCV onto Stellar rails. We’ve also shipped multiple Stellar Community Fund–funded deliverables (e.g., Loam, Moonlight, Scaffold Stellar), and we actively support ecosystem governance and developer education (mostly through hackathons).
Turbolong is a live, production deployment, not a concept pitch. Traction falls into four verifiable categories:
Mainnet Blend-Etherfuse / Fixed and Yieldblox pools integration: all three are referenced in frontend/src/blend.ts, scripts/mainnet_loop.ts, and the Rust simulator.
Soroban WASM strategy contract (contracts/strategies/blend_leverage) covered by an integration suite including health-factor invariants and pool-state parity assertions against a live mainnet fork via soroban-ledger-snapshot-source-tx.
Approved for Drips Stellar Wave (drips.network/wave/stellar), the only Blend-native leverage application on the Drips contributor list. 98 structured, labelled contributor issues are open at https://github.com/Dgetsylver/TurboLong/issues, sized S/M/L and labelled across 11 areas for the monthly wave cycles.
Internal UX research: 100-persona audit grounding the roadmap in real DeFi-user, beginner-user, and LatAm-saver segments.
Security posture: Turbolong-specific profitability and attack-cost analysis (profitability_analysis.md) showing the four structural attack vectors against recursive-loop protocols are economically non-viable on Stellar thanks to the absence of flash loans and the backstop-capped rate spread. Two independent vulnerability reports published.
Active composition with four Integration-List building blocks in production today (Blend v2, DeFindex, Soroswap, Stellar Wallets Kit), which is both the foundation this application builds on and the reason the application's additions (Stellar Broker, Aquarius) compose cleanly.
Alerts service deployed on Cloudflare Workers (alerts/) with D1 persistence and a Resend-backed verified-email flow, the only public leverage-alerts service on Stellar.
Designed around Etherfuse stablebonds, giving Turbolong a natural regulatory story: government-bond-backed stable assets amplified by Stellar rails.
1. DeFindex — full mainnet vault deployment
Description: Deploy the blend_leverage strategy on mainnet via DeFindex across four assets (USDC, CETES, USTRY, XLM).
How to measure completion: Four mainnet DeFindex vaults live (one per asset); end-to-end deposit → loop → withdraw flow verified on Stellar Expert for each; frontend placeholder removed; DeFindex team co-signs the deployment.
Budget: $7,000
2. DeFindex — SEP-41 transferable receipt token
Description: Mint a SEP-41 receipt token representing vault-share ownership. Needed as a prerequisite for the Aquarius listing in Tranche #3. Includes storage schema, transfer semantics, and a migration path from existing non-tokenized positions.
Measure of completion: Receipt token contract deployed on mainnet; transfer semantics verified by unit + integration tests; existing depositors migrated without position loss.
Budget: $4,000
3. DeFindex — strategy versioning & migration helper
Description: Add a migrate() entry point so deploying a v2 strategy doesn't force users to exit and re-enter. Per-user, signature-gated migration preserving HF and balances within rounding.
How to measure completion: Migration tested against live pool-state fixtures; HF and balance parity checked within 1e-7; documented runbook for future strategy upgrades.
Budget: $2,000
4. Stellar Wallets Kit — mobile deep-linking + Ledger
Description: Extend kit usage so Lobstr, xBull, and Hana users can sign from a mobile deep-link without copy-paste. Add Ledger hardware-wallet support via the kit adapters. All signing flows remain kit-native.
How to measure completion: E2E tests passing for all 5 supported wallets on both classic and Soroban ops; mobile deep-link flow verified on iOS and Android builds of Lobstr and xBull.
Budget: $5,000
5. Foundational CI / supply-chain hygiene
Description: GitHub Actions workflows for frontend, Rust, and Worker (lint, clippy, fmt, tests). Enable Dependabot and cargo-audit. Commit gitleaks config and pre-commit hooks. Unblocks reliable contribution across the remaining tranches and the Drips wave.
How to measure completion: CI green on main; Dependabot opening weekly PRs; cargo-audit passing; gitleaks blocks a deliberately planted secret in a test PR.
Budget: $2,000
1. Stellar Broker — primary swap routing
Description: Integrate Stellar Broker as the primary swap router for BLND harvest and any intra-vault swaps. Route chosen per-tx based on quoted output; Soroswap retained as fallback when Broker's quote is worse or unavailable. Slippage protection wired at both paths.
How to measure completion: ≥ 50 mainnet harvests with both Broker and Soroswap quotes logged; per-tx best-route selection verifiable against quoted outputs; fallback path triggers correctly when the primary fails or returns a worse quote; post-tranche analysis report documents Broker-vs-Soroswap win rate and average uplift. Acceptance is on the routing logic and the data set, not on which router wins more often.
Budget: $10,000
2. DeFindex — partial-unwind liquidation protection
Description: If HF drops into an orange zone (configurable threshold), the keeper unwinds just enough loops to restore target HF rather than forcing a full close. Minimises user cost and pool impact.
How to measure completion: Algorithm verified against fixture set covering all degenerate cases; dry-run simulation matches on-chain result within rounding.
Budget: $8,000
3. DeFindex — auto-rebalance keeper
Description: Keeper-authorised entry point that unwinds N loops to restore target when HF drops below the configured min_hf. Rate-limited; emits events with before/after HF and loops unwound. Unit tests cover edge cases (locked reserves, already at floor).
How to measure completion: Keeper role authenticated; event emitted on every rebalance; 100+ simulated rebalances in testnet; at least one live mainnet rebalance during tranche.
Budget: $6,000
4. Post-loop projected rates
Description: Simulate the user's own supply/borrow impact on pool utilisation before they sign. Display projected post-loop net APY instead of the pre-loop snapshot. Math verified against the Rust simulator.
How to measure completion: Parity tests between frontend/src/blend.ts projectRates and the Rust simulator passing across ≥ 20 fixtures spanning all three IR-curve kinks.
Budget: $3,000
5. Historical APY storage (enables Aquarius comparison in T3)
Description: Extend the alerts service: (a) add a rate_snapshots D1 table written every 15 minutes by the existing cron, with public GET endpoint and 365-day retention; (b) extend the subscriptions table with hf_threshold, alert_type ('apy' | 'hf' | 'liquidation' | 'rate_spike'), and last_fired_at columns to support health-factor and liquidation-imminent alert channels. Same rate_snapshots table powers the 24h/7d delta arrows added in Tranche #3.
How to measure completion: ≥ 30 days of rate-snapshot data accumulated by end of tranche; public endpoint serving paginated JSON; pruning job tested; HF and liquidation-imminent alert channels live with at least one alert fired in production.
Budget: $3,000
1. Aquarius — rate comparison and 'Best Rate' UI
Description: Query Aquarius' liquidity and routing protocol for live XLM/USDC and asset-pair rates across the aggregated Stellar DEX surface. Surface a 'Best Rate' badge in the Compare Pools view and drive 24h/7d rate-trend arrows from the T2 historical snapshot table.
How to measure completion: Live Aquarius rate data visible in UI for ≥ 5 asset pairs; badge correctness verified against Aquarius' own routing API; fallback path documented when Aquarius is unreachable.
Budget: $9,000
2. Aquarius — SEP-41 receipt token listing
Description: List the DeFindex vault-share receipt token (minted in T1) on Aquarius so depositors can exit positions without going through the full DeFindex unwind path.
How to measure completion: Receipt token tradable on Aquarius; at least 5 external trades executed against the market during tranche; trading UX integrated into the Turbolong position-management view.
Budget: $10,000
3. Compare Pools UI — Aquarius + historical charts
Description: Build out the Compare Pools view (issue A9) with Aquarius data on one axis and the T2 historical rate table on the other. 7d/30d/1y APY charts per asset; rate-trend arrows; cross-pool net-APY ranking at multiple leverage brackets.
How to measure completion: Compare Pools view deployed; charts render in < 100 ms; screenshots included in mainnet-launch blog post; view accessible without a wallet connection.
Budget: $6,000
4. Testing
Description: Run a structured internal testing programme across sessions weighted toward LatAm and DeFi-native users, covering the full deposit → monitor → exit flow. Findings triaged by severity and tracked as labelled GitHub issues.
How to measure completion: Structured test plan; completed sessions with recorded task-success scores; findings report.
Budget: $8,000
5. Mainnet launch — fully composed product
Description: all Tranche #1-#3 surfaces live simultaneously on mainnet. Public status page, updated onboarding tour, EN/ES/PT-BR strings for the core trade flow, post-launch runbook, formal SCF #43 completion report.
How to measure completion: All composed surfaces green on mainnet for 14 consecutive days before launch announcement; status page live.
Budget: $6,000
Engineering and business background with hands-on Web3 operator experience. Leads strategy, partnerships, and execution at The Aha Company, turning complex blockchain systems into shippable products and scalable business lines.
Connect → LinkedIn
20 years in tech, with multi-year Web3 and frontier AI experience at Aragon (Ethereum DAOs), Lilypad Network (distributed compute on Arbitrum, as CTO), and Scale AI (data capture infrastructure for computer-use AI models). Deep expertise in Web3 product engineering and AI infrastructure, with experience across Ethereum, Arbitrum, IPFS, and multi-chain wallet systems, plus a track record leading cross-functional teams of 6 to 20+ engineers.
13+ years in tech, with deep experience leading engineering across Web backend and infrastructure for Series A/B startups and large enterprises in EdTech and Blockchain. Joined The Aha Company in October 2025 following 4+ years at Cheesecake Labs where he scaled from Lead Engineer to Engineering Manager — running a 25-person remote team, launching an internal AI innovation practice, and shipping across dedicated, full-product, and staff augmentation engagements. Strong systems chops across Node.js, TypeScript, GoLang, PHP, hexagonal architecture, and IaC, paired with proven leadership in technical strategy, mentoring, and cross-functional delivery.
Connect → Github | Linkedin
4 years in tech, with 3+ years specializing in blockchain; first as an IT transformation / blockchain consultant at BearingPoint in Paris, advising on enterprise blockchain initiatives, now operating independently from Geneva since 2024. Strong frontend foundation in React, React Native, and JavaScript applied to Web3 client work. EPITA graduate with international exposure via California State University, Monterey Bay.
Connect → Linkedin
Enzo Soyer
No other submissions.