ChainSente: A Cardano Wallet for DeFi and Mobile Money

[Proposal setup] Proposal title

Please provide your proposal title

ChainSente: A Cardano Wallet for DeFi and Mobile Money

[Proposal Summary] Budget Information

Enter the amount of funding you are requesting in ADA

48000

[Proposal Summary] Time

Please specify how many months you expect your project to last

6

[Proposal Summary] Translation Information

Please indicate if your proposal has been auto-translated

No

Original Language

en

[Proposal Summary] Problem Statement

What is the problem you want to solve?

Cardano has wallets, not credit. Without a safe mobile way to borrow against ADA, people sell to fiat for daily needs paying fees and leaving Cardano. ChainSente keeps value and trust here.

[Proposal Summary] Supporting Documentation

Supporting links

• https://x.com/chainsente
,
• https://github.com/Nazrawi15/ChainSente
,
• https://medium.com/@ChainSente
,
• https://docs.google.com/document/d/1wg63uwga14APBBDh2DFyLzrBP5XLlqRwW_lsPKP9668/edit?usp=sharing

[Proposal Summary] Project Dependencies

Does your project have any dependencies on other organizations, technical or otherwise?

No

Describe any dependencies or write 'No dependencies'

No dependencies

[Proposal Summary] Project Open Source

Will your project's outputs be fully open source?

Yes

License and Additional Information

Apache 2.0 — all code, tests, and documentation will be fully open source for free use, modification, and redistribution with attribution. This covers the Plutus or Aiken contracts, React Native wallet, oracle simulator, and CLI tools, with all artifacts published in a public GitHub repo for community reuse and verification.

[Theme Selection] Theme

Please choose the most relevant theme and tag related to the outcomes of your proposal.

Lending

[Campaign Category] Category Questions

Describe what makes your idea innovative compared to what has been previously funded (whether by you or others).

ChainSente is a new way to test lending on Cardano for real people in Africa. Users lock ADA; we price it in USDT and show how much they can borrow in UGX, with clear safety limits and auto-liquidation if prices fall. It ships with a simple phone wallet to send or receive money on testnet and clear screens that simulate withdrawals to Airtel and MTN mobile money now built to plug into real rails later. We open source the contracts and a price shock tester so anyone can verify the math. No hype just a safe MVP.

Describe what your prototype or MVP will demonstrate, and where it can be accessed.

By project end, ChainSente will have a working MVP on the Cardano testnet. You’ll get smart contracts live on testnet (lock, borrow, repay, liquidation) with public contract IDs (script hashes) and addresses; an Android app (apk) to create a test wallet, lock ADA, see the UGX loan amount, repay, and view each step on a blockchain explorer; plus a small CLI and clear guides on GitHub. Access: GitHub repo, apk download, and testnet explorer links. Community: install the app, run the flows, try price shock demos, report bugs, and fill a short survey.

Describe realistic measures of success, ideally with on-chain metrics.

By project end we target: 150–200 testnet wallets, 120–150 full lending cycles (lock → borrow → repay), and 600–800 total contract calls across our scripts. At least 200 tx to the lock script and 100 repays executed. Zero invariant breaches; liquidation events recorded on-chain in ≥10 price shock runs. P95 confirmation (from the explorer). We’ll publish script hashes, tx IDs and a small dashboard so anyone can audit every number.

[Your Project and Solution] Solution

Please describe your proposed solution and how it addresses the problem

The problem we’re solving

In East Africa, small cash needs (stock, transport, school, emergencies) are handled through mobile money. Crypto lending on Cardano isn’t designed for these users: tools assume crypto-native behavior, unstable UX, and unclear risk. People don’t need a new coin; they need a safe, simple way to borrow against crypto they already hold, with guardrails and clear outcomes—before anyone talks about real Airtel/MTN rails.

Our approach and why this design

We ship a testnet only, non-custodial MVP that proves the lending loop, safety rules, and usability—without partners, fiat custody, or regulatory exposure. Users lock ADA, we price it in USDT and convert to UGX for a displayed loan amount. We enforce loan-to-value (LTV) safety bands and auto-liquidation if price falls, and we make every rule auditable on-chain. Withdraw to Airtel/MTN is simulated in-app during MVP to validate UX; live rails are a Phase-2 decision based on evidence.

This scope is deliberate: it removes dependencies, accelerates learning, and lets the community verify real usage through script hashes and tx IDs.

What the MVP demonstrates

1. Smart contracts (Plutus/Aiken) for lock → borrow → repay → liquidation, deployed on Cardano testnet.
2. Oracle simulator: a signed price-feed UTXO (ADA/USDT, USDT/UGX) used as a reference input so anyone can replay price changes and trigger the same outcomes.
3. Invariant enforcement (safety math) proven by property tests and on-chain events:
   • Collateral value ≥ required margin, else liquidation.
   • Repay unlocks exactly principal; no unauthorized state transitions.
4. Mobile wallet (React Native) to send/receive ADA (testnet), create/import wallet, view history, request a UGX-priced loan, and simulate Airtel/MTN withdrawal via guided screens.
5. Price-shock simulator + CLI: demonstrate partial liquidation restoring LTV to safe bounds.

How it works

• User locks C ADA into the contract.
• Oracle publishes $P_{A/U}$ (ADA→USDT) and $P_{U/G}$ (USDT→UGX).
• App shows a borrow limit using $L = C \times P_{A/U} \times P_{U/G} \times LTV_0$ (e.g., 50%).
• We maintain a safety band $m = 10\text{–}25%$. If live LTV ≥ $1-m$ (e.g., 75% at m=25%), liquidation fires; we partially liquidate just enough ADA to restore target LTV.
• All steps emit on-chain events so results are verifiable.

Who we engage

• Pilot users in Uganda/Kenya who are already comfortable with mobile money and curious about crypto—micro-retailers, drivers, students.
• Cardano community testers who install the APK, run the flows, and file GitHub issues.
• Local devs and advocates who want open, reusable lending primitives and a wallet template for Africa-first UX.

We support English, Luganda, and Swahili, low end Androids, clear error copy, and short tutorials. Two workshops/seminars (with refreshments, T-shirts, and stationery) recruit and train testers and gather feedback.

,

Why Cardano

UTXO + Plutus gives deterministic state, low fees on testnet, and clean audit trails via reference inputs and datum-driven state machines. That’s ideal for transparent risk controls and replayable tests—critical before touching any real payment rails.

What is in and what is out (to keep trust)

In scope (MVP/testnet):

• Contracts, oracle simulator, liquidation logic, wallet app, CLI, workshops, public dashboards, and reports.

Out of scope (Phase 2 decision):

• Real Airtel/MTN integrations, fiat custody, KYC, or mainnet launch. We will only consider these after KPIs are met and risks documented.

How we prove impact

• On-chain: public script hashes, testnet addresses, and a dashboard showing:
  • 150–200 wallets created (testnet)
  • 120–150 full loan cycles (lock→borrow→repay)
  • 600–800 total contract calls
  • 10+ liquidation events from price-shock runs
  • Zero invariant breaches, P95 confirmation <60s
• Off-chain but auditable: APK link, CLI usage, GitHub tests/coverage, anonymized pilot CSV (task success, times, error rates), and SUS ≥70.
• Community verification: every metric links to tx IDs and reproducible scripts.

Risk and mitigation

• Price-feed trust: MVP uses a signed mock oracle with public keys; anyone can reproduce price paths.
• User shock at liquidation: we add alerts, grace tips in the simulator, and cap liquidation to “just enough” to restore safety.
• Scope creep: hard “out-of-scope” line; any third-party API = Phase 2.
• Talent bottlenecks: pre-vetted Plutus and RN contractors; QA gates in CI to prevent regressions.

Reuse and community value

Everything is open-source (contracts, RN wallet template, oracle and price-shock tools, test harness). Teams can fork the code to build regional products or plug different collateral/price feeds. Our docs include threat models, failure-mode playbooks, and a step-by-step “how to verify our results” guide for reviewers.

Why this will work

• It solves the smallest real problem first: safe borrowing UX with transparent rules.
• It’s verifiable (hashes, tx IDs, tests) and inclusive (languages, low-end devices).
• It creates reusable building blocks for African DeFi without overpromising on partners or regulation.
• It gives Catalyst a clean go/no-go decision for Phase 2 based on data, not hype..

Summary

ChainSente delivers a testnet-only, non-custodial lending MVP on Cardano: lock ADA, see a UGX-priced borrow amount, and borrow/repay with LTV safety bands (10–25%) and auto-liquidation—all audited on-chain. A mobile wallet and oracle simulator make it usable and provable; workshops recruit real users. We measure success by tx IDs, invariant tests, and pilot data, then publish everything so the community can verify—and decide if live rails are warranted.

[Your Project and Solution] Impact

Please define the positive impact your project will have on the wider Cardano community

What value ChainSente brings to Cardano

• DeFi primitives you can reuse: Open-source Aiken/Plutus contracts for ADA-collateralized lending with LTV bands and partial liquidation, plus an oracle/price-shock simulator. These reduce time-to-prototype for any lending product on Cardano.
• Africa-first UX patterns: A React-Native wallet template, multilingual flows (EN/LG/SW), and a “mock Airtel/MTN” experience that others can adapt for real rails later.
• Proof, not promises: Testnet addresses, script hashes, and reproducible CLI runs create a transparent record of what works, what doesn’t, and why—useful for teams, auditors, and educators.

Who benefits

• Builders: Copy/paste lending state machines, invariants, and device-tested mobile UX.
• Researchers and auditors: A public data set of on-chain events + price-shock runs to study risk controls.
• Community ; Practical patterns for onboarding non-crypto users in East Africa without over-promising on partners or regulation.

How we measure impact (quantitative plus qualitative)

On-chain (verifiable by anyone)

• 150–200 testnet wallets created.
• 120–150 full loan cycles (lock → borrow → repay).
• 600–800 total contract calls across our scripts.
• ≥200 lock tx; ≥100 repay tx recorded on Cardano testnet.
• 10+ liquidation events in controlled price-shock tests.
• 0 invariant breaches (enforced by property tests + on-chain checks).
• P95 confirmation (timed from explorer).

Off-chain but auditable

• Usability: System Usability Scale (SUS) ≥70; task-success ≥80% within three tries.
• Reliability: Crash-free session P90 ≥20 min across 10+ test devices.
• Field notes: Anonymized CSV (errors, completion times, common failure points), plus short interviews from two workshops.

Evidence andtooling

• Public script hashes, testnet addresses, tx IDs.
• A small web dashboard that reads the explorer and shows counts in real time.
• CLI simulator to replay price paths and reproduce liquidation events on any machine.

How we share outputs

• Code and docs: GitHub (contracts, RN wallet template, oracle/CLI, test harness, threat model, failure-mode playbook). License: Apache-2.0 (or MIT) for easy reuse.
• Releases: APK download + step-by-step “Verify our results” guide (with explorer links).
• Reports: End of milestone posts with KPIs, tx IDs, and lessons learned.
• Community channels: Cardano Forum write-ups, X/Twitter dev threads, short demo videos, and a post in Cardano Africa / builders spaces.
• Workshops: Two open sessions (with refreshments/T-shirts/stationery) to train testers; slides and recordings posted.

What success looks like for the community

• Teams can fork our contracts and wallet flows to launch region-specific pilots faster.
• New devs learn from clean, tested Aiken examples instead of starting from scratch.
• Catalyst gains a data-backed view of whether ADA-collateralized lending with partial liquidation is understandable to first-time users—before anyone touches real mobile-money rails.
• A clear Phase-2 decision (go/no-go for partners and mainnet) based on metrics, not hype.

If results are mixed (impact still holds)

• Even if liquidation UX proves confusing or margins need tuning, we publish the negative results and fixes. That saves other teams months of trial and error and reduces user risk across the ecosystem.

Inclusion, ethics, and privacy

• No real KYC or fiat in MVP; pilots run on testnet.
• In workshops we use consent forms, minimize personal data, and publish anonymized findings only.
• UX is tested on low-end Android devices and in English, Luganda, and Swahili.

Long-term path

• If KPIs are met, we propose Phase-2: one vetted payment aggregator sandbox, clearer liquidation alerts, and expanded languages. If not, the community still keeps the libraries, templates, data, and playbooks.

Impact KPIs (summary, all public)

150–200 wallets; 120–150 full cycles; 600–800 contract calls; ≥200 locks; ≥100 repays; 10+ liquidation demos; 0 invariant breaches; P95 and lt 60s; SUS ≥70; task success ≥80%; P90 crash-free ≥20 min.Where to verify: GitHub (code, CLI, guides), testnet explorer (hashes + tx IDs), dashboard screenshots, workshop reports.

Bottom line: ChainSente turns lending on Cardano into something you can run, measure, and reuse—with open artifacts that help the next wave of builders move faster and safer

[Your Project and Solution] Capabilities & Feasibility

What is your capability to deliver your project with high levels of trust and accountability? How do you intend to validate if your approach is feasible?

Why we are suited to deliver

• Tight, dependency-free scope. ChainSente ships a testnet-only, non-custodial MVP: lock ADA → price in USDT → display UGX loan limit → borrow/repay with LTV safety bands (10–25%) and auto-liquidation. No real Airtel/MTN rails, no KYC, no custodial balances. This keeps risk low and delivery predictable.

• Team roles mapped to milestones.

  • Plutus/Aiken dev (M2): state machine, liquidation, property tests, CLI simulator.

  • React-Native dev (M3): wallet (create/import, send/receive testnet ADA), lending flows, i18n.

  • UX/Field (M1 & M4): Figma flows, workshops, SUS/task testing, pilot analytics.

  • QA/Automation (M2–M4): CI, unit/E2E tests, device matrix, release gates.

  • PM & Docs (all): scope control, reports, hashes/tx IDs, risk register.

• East-Africa context. We design for low-end Android, short sessions, and English/Luganda/Swahili. Two workshops (with refreshments, T-shirts, stationery) bring real users into the loop early.

How we validate feasibility

• On-chain truth: Contracts live on Cardano testnet with public script hashes and addresses. Every borrow/repay/liquidation emits events that anyone can audit on a testnet explorer.

• Safety proven, not asserted:

  • Invariants: (1) collateral value ≥ required margin; (2) repay unlocks exactly principal; (3) no unauthorized state transitions.

  • Property-based tests and a price-shock CLI that replays ADA→USDT and USDT→UGX moves to trigger liquidations and partial re-collateralization.

• Usability proven with data: Two workshops + pilot cohort; targets: SUS ≥70, task success ≥80% within 3 tries, crash-free P90 ≥20 min, P95 confirmation <60s (from explorer).

• KPIs (end of MVP): 150–200 testnet wallets; 120–150 full lending cycles; 600–800 total contract calls; 10+ liquidation demos; 0 invariant breaches.

Process that earns trust & accountability

• Milestone gates with pass/fail evidence

  • M1—Specs (Mo1): diagrams, threat model, LTV math, oracle-sim design. Gate: 2 external reviews resolved in GitHub; spec PR merged.

  • M2—Contracts (Mo2–3): scripts + unit/property tests + CLI; script hashes posted. Gate: CI ≥95% tests pass; invariants proven; demo tx IDs published.

  • M3—Wallet (Mo4–5): RN app (create/import, send/receive, history, i18n); Gate: ≥30 testnet tx across ≥10 devices; crash-free P90 ≥20 min.

  • M4—Pilots & Workshops (Mo6): mock Airtel/MTN flows + analytics; Gate: ≥40 unique users complete flows; SUS ≥70; task success ≥80%; anonymized CSV + dashboard posted.

• Public artefacts, every step:

  • GitHub (Apache-2.0): contracts, RN app, oracle & price-shock tools, CI configs, coverage, threat model, failure-mode playbook, how-to verify.

  • Explorer links: script hashes, addresses, tx IDs attached in reports.

  • Monthly demos: short videos + changelogs.

• Risk management (written and enforced):

  • Scope creep: any third-party API or fiat rails = Phase 2 only.

  • Talent risk: pre-vet 2 Plutus and 2 RN candidates; keep one backup; start with time-boxed trial tasks.

  • Security & correctness: property tests + mutation tests; peer review on every contract PR; “two-reviewer” rule before merges.

  • Device fragmentation: test matrix of 10+ Android devices/versions; fall back to APK side-load if Play distribution is slow.

  • Liquidation UX shock: alerts in app, clear explanations, and partial liquidation (only “just enough” ADA) to restore LTV.

Financial controls (how we handle funds properly)

• Milestone-based disbursement. Funds are drawn only after passing the gate criteria above; evidence (hashes/tx, CI screenshots, APK) is posted before each draw.

• 2-of-3 multisig wallet for treasury (PM + one technical lead + one community advisor).

• Budget discipline: hours × rates are tracked in a shared ledger (role, milestone, deliverable link, tx ref). Contingency = 10%; use requires a short public note explaining deviation and link to the fix.

• Procurement policy: ≥500 ADA requires 2 quotes (devices, venue, merch). Receipts and invoices are uploaded to the repo’s /reports/finance folder with tx references.

• No custody of user funds: MVP runs on testnet only; no user deposits, no fiat handling.

Why this approach is feasible (design choices that remove risk)

• No external dependencies. No payment aggregators, no KYC vendors, no price-oracle reliance beyond a signed mock feed you can reproduce locally.

• Smallest useful surface. One collateral (ADA), a single LTV policy band, clear liquidation math—easier to test and reason about than multi-asset designs.

• Deterministic audit trail. UTXO + datum-driven state machine + reference inputs make every state transition replayable and explainable.

What we’ve already prepared (to hit the ground running)

• Repo skeleton with CI, docs scaffolding, issue templates, and checklists for each gate.

• Pilot plan & materials: consent script, feedback forms, SUS survey, facilitator guide; venues pre-scoped with costs (workshops include refreshments, T-shirts, stationery).

• Device pool plan: two mid-range Android test phones, plus community testers’ devices during workshops.

What the community can verify (and how)

• Scripts & addresses: see /reports/M2/ for script hashes and testnet addresses; click explorer links.

• Invariants & tests: run npm run test (wallet) and aiken test (contracts); property tests print invariant stats.

• Price-shock simulator: run chainsente-cli shock --path sample.json to reproduce a liquidation; compare tx IDs with our report.

• User results: open the dashboard to see wallet counts, contract calls, and liquidation events pulled from the explorer; download the anonymized CSV from /pilot/.

If something goes wrong (our fallback paths)

• Schedule slip in contracts: de-scope optional UI polish; keep invariants and test coverage intact.

• RN delays: ship Android APK first; iOS moves to Phase 2.

• Workshop turnout low: add one remote testing day; increase stipends/airtime within the pilot budget line (documented).

• Invariant breach found: pause new features, fix root cause, publish a post-mortem and rerun the property tests; do not proceed to next milestone until green.

---

Bottom line: We operate with "small, testable milestones*, each with objective pass/fail gates, and we publish hashes, tx IDs, code, tests, and financial records so any voter or reviewer can check our work. By removing external dependencies, keeping the design simple and safe, and enforcing process (CI, reviews, multisig, receipts, and public reports), we make delivery feasible, auditable, and accountable from day one.

[Milestones] Project Milestones

Milestone Title

Architecture, Risk Specs and UX Flows

Milestone Outputs

• We produce the lending control policy (LTV bands, safety buffer, partial-liquidation sizing), a signed mock oracle spec (freshness and monotonic time rules), a concise threat model with FMEA/HAZOP risk ranking, and complete Figma flows for onboarding, lock → borrow → repay, and error states. We also add a glossary, an ADR (decision log), and a Monte Carlo plan to validate price-shock behavior.

Acceptance Criteria

• Two external reviewers read the specs and we resolve all comments; the spec PR is merged. All invariants (e.g., “repay unlocks principal exactly,” “collateral ≥ requirement,” “oracle time monotonic”) are listed and mapped to tests. The control-policy table is filled with initial parameters, and a small Monte Carlo sample proves the policy avoids over-liquidation under typical shocks.
• FIGMA PLAN: https://docs.google.com/document/d/1wg63uwga14APBBDh2DFyLzrBP5XLlqRwW_lsPKP9668/edit?usp=sharing

Evidence of Completion

• /docs/ specs with commit hash and tag, review issue links, the Figma share, a checklist mapping each invariant ↔ planned test, the control-policy table, a brief Monte Carlo summary (CSV + plots), the risk register with RPN scores, ADR entries explaining choices, and a one-page glossary. Reviewer sign-off (names, dates) is included.

Delivery Month

1

Cost

4950

Progress

10 %

Milestone Title

Contracts Deployed on Testnet Plus Price Shock CLI

Milestone Outputs

• We will ship Plutus sate machine contracts for lock, borrow, repay, and partial liquidation; a full unit + property-test suite for the invariants; and a CLI simulator with sample ADA→USDT→local price paths so anyone can replay shocks. Deployed script hashes and testnet addresses are published with a concise README and test vectors.

Acceptance Criteria

• Continuous Integration shows ≥95% tests passing and 0 invariant breaches across stress sets. We post explorer tx IDs demonstrating lock, borrow, repay, and liquidation. Contracts reject stale or out-of-order oracle updates. Code is peer-reviewed, coverage reported, and mutation/static-analysis checks are green. Gas/fee budget fits our constraints on testnet.

Evidence of Completion

• Public repo with contracts/tests, green CI + coverage badges, CLI docs and vectors, a tagged release, and explorer links for each demo flow. We include script hashes, testnet addresses, test logs, and a short demo video of the CLI triggering a clean partial liquidation that restores LTV to target + buffer without over-selling collateral.

Delivery Month

3

Cost

21230

Progress

30 %

Milestone Title

Android Wallet Core (Testnet)

Milestone Outputs

• An installable Android APK that lets users create/import a wallet, send/receive testnet ADA, run the lending steps (lock, see safe limit, borrow, repay), and view transaction history. Release includes secure key storage, QR receive, crisp error states, English copy, checksum for file integrity, device-matrix notes, and a short “Getting Started” guide.

Acceptance Criteria

• We execute 30+ real testnet transactions on 10+ different devices; P90 crash-free time ≥20 minutes as measured by Crashlytics; and P95 confirmation <60s based on explorer timing. Basic performance (cold start, navigation) remains smooth, and no critical accessibility or error-handling gaps remain in the core flows.

Evidence of Completion

• Published APK + checksum, Crashlytics stability report, the device-matrix results, explorer links for every demo transaction (lock/borrow/repay/liquidation), annotated screenshots of key screens, and release notes detailing changes and known limits. A short screencast shows the full end-to-end flow on a mid-range Android.

Delivery Month

5

Cost

13100

Progress

80 %

Milestone Title

Pilot and Workshops plus Public Reporting

Milestone Outputs

• Two hands-on workshops (run by Nasser and Salm) using simulated mobile-money steps; a pilot cohort that completes the end-to-end flows; and a public dashboard + anonymized CSV with usage and usability data. We publish a final report summarizing KPIs, key lessons, and a go/no-go recommendation for Phase 2 (consider real rails only if evidence supports it).

Acceptance Criteria

• At least 40 unique users complete the core flows; SUS ≥70 and task success ≥80% within three tries; we run 10+ clean liquidation demos via price-shock runs; and tests continue to show zero invariant breaches. We also capture rural/urban split and top three UX fixes implemented before close.

Evidence of Completion

• Workshop packs (slides, facilitator notes), photos (no personal data), dashboard screenshots, the anonymized CSV, a short demo video, explorer tx IDs for pilot runs, and the final report with KPIs, lessons, and Phase-2 decision. Links are centralized in /reports/M4/ with commit tags for traceability

Delivery Month

6

Cost

8700

Progress

100 %

[Final Pitch] Budget & Costs

Please provide a cost breakdown of the proposed work and resources

1. People (hours × rate = ADA) — 36,600 ADA (76%)

• Plutus smart-contract engineer — builds state machine + tests + CLI 320h × 55 = 17,600 ADA

•     M1 Support (specs & data model) — 1,600 ADA 
  

•     M2a State machine + unit tests (PRs merged) — 7,000 ADA
  

•     M2b Property tests for invariants (CI green, coverage posted) — 5,000 ADA
  

•     M2c Testnet deploy + CLI + demo tx IDs — 4,000 ADA
  

• React-Native mobile engineer — Android app (wallet + lending flows) 280h × 34 = 9,520 ADA
• UX & Field Research — Figma, user testing, workshops materials 170h × 20 = 3,400 ADA
• QA/Automation and DevOps — CI, device matrix, Crashlytics, releases 160h × 20 = 3,200 ADA
• PM and Docs — planning, reports, finance ledger, community updates160h × 18 = 2,880 ADA

Why this is lean: most ADA funds people who ship code, tests, and reports—not marketing fluff.

2. Non-people resources — 7,036 ADA (15%)

• Infra and; tools — CI minutes (buffer), domain/TLS, tiny demo VM, test SIM/data, two mid-range Android test phones → 1,400 ADA

* CI is free for public repos (we still keep a buffer).

* Crashlytics crash reporting is no-cost on Firebase.

* Small demo server (e.g., Render Basic-256MB ≈ $6–7/mo) → covered in the 1,400 ADA infra pot.

• Community pilot — stipends/airtime/transport for 40–60 testers → 2,200 ADA
• Workshops and seminars (×2) — venue, refreshments, water, T-shirts, stationery/printing, PA/projector, photography → 3,436 ADA

Why this is lean: we avoid paid licenses and heavy hosting; we rely on open-source + public CI + a tiny VM.

3. Contingency — 4,364 ADA (9%)

• Held for unexpected fixes (contract refactors, device replacements, extra pilot day). Any use is explained in a public note and linked to the fix. If unused, it remains unspent.

4. Per-milestone budget mapping (what you get for each payment)

• M1 — 4,950 ADA: control policy (LTV bands + buffer), oracle simulator spec, FMEA/HAZOP, Figma flows; /docs + reviewer sign-off.
• M2 — 21,233 ADA: deployed contracts, property tests, price-shock CLI, script hashes & tx IDs on testnet.
• M3 — 13,111 ADA: Android APK; 30+ testnet tx on 10+ devices; Crashlytics report; P90 ≥ 20m crash-free, P95 and lt; 60s confirmation.
• M4 — 8,706 ADA: two workshops/seminars, pilot dashboard, anonymized CSV, final report with go/no-go for Phase 2.

Total: 4,950 + 21,233 + 13,111 + 8,706 = 48,000 ADA

5. Third-party products and services (what, why, and cost stance)

• Firebase Crashlytics (free) — crash reporting and stability metrics. No fee expected at MVP scale; fits into the Firebase no-cost tier.
• GitHub Actions (free for public repos) — CI/CD with tests and coverage; we budget only a small buffer under Infra in case we flip any private runs.
• Render (or similar) tiny VM — hosts the demo dashboard/CLI API during pilot; priced around $6–7/month; covered within Infra.
• Domains/TLS — one domain and certificates (low cost, included in Infra).
• Accounting/legal/code audit — not required at MVP/testnet scope; any ad-hoc professional help, if needed, comes from contingency and is documented.

6. Publicity, community, documentation (often forgotten — we included them)

• Workshops & seminars (venues, refreshments, T-shirts, stationery) are fully budgeted under 3,436 ADA.
• Community engagement — posts, demo videos, and reports are part of PM & Docs hours.
• Reporting back — we publish APK + checksum, script hashes, tx IDs, dashboards, anonymized CSV, and finance ledger in the repo.

7. How we manage and prove proper use of funds

• 2-of-3 multisig treasury (Nasser, Salm, Semir). No payout until public evidence is posted (hashes, addresses, tx IDs, CI/coverage, APK checksum, device-matrix, pilot CSV).
• /reports/finance/ ledger with payouts + tx IDs and receipts.
• Open-source (Apache-2.0): all code, tests, and reports with explorer links for audit.

summary (why this is valuable for Cardano)

• Spend is focused on build and; proof, not marketing.
• We use free tiers where possible (Crashlytics, public CI) and a tiny server for demos.
• Workshops/seminars are funded to get real feedback and a reusable dataset.
• Everything is public and verifiable (script hashes, tx IDs, dashboards).
• Result: a reusable DeFi MVP for Cardano with measurable on-chain impact and clear governance—for 48,000 ADA, fixed.

[Final Pitch] Value for Money

How does the cost of the project represent value for the Cardano ecosystem?

What you get for 48,000 ADA

A testable, open, on-chain MVP: audited lending contracts, a price-shock simulator, a working Android testnet wallet, and a public dataset + report. All code and results are open so other Cardano teams can reuse them.

Built for the “Concepts” track

We spend to ship → measure → publish. The goal is a real prototype with real testnet transactions and clear learning for the ecosystem—not a slide deck.

Where the ADA goes (lean and focused)

• People who ship code & tests (Plutus/Aiken, RN, QA, PM): 36,600 ADA → 76%
• Infra + pilots + workshops: 7,036 ADA → 15%
• Contingency (unexpected fixes): 4,364 ADA → 9% Every ADA buys a concrete artifact: contracts, test harness, APK, dashboards, and explorer tx IDs you can click.

Why this is cost-effective

• Local execution, global quality. We use Uganda-based talent for UX/QA/DevOps and focused contractors for Plutus/RN. Lower local rates + strict quality gates (CI, coverage, invariants) = more verified outputs per ADA.
• Testnet first = cheaper learning. No fiat rails, no custody, no partner costs in MVP. We prove the lending loop and safety rules before paying for integrations.
• Open source = compounding ROI. Apache-2.0 licensing means any Cardano team can fork the Aiken state machine, invariant tests, price shock CLI, and RN wallet flows reducing future grant spend.

Cost → proof (per milestone)

• M1 – 4,950 ADA: control policy (LTV bands + buffer), oracle simulator spec, risk review (FMEA/HAZOP), Figma flows. Prevents expensive rewrites later.
• M2 – 21,233 ADA: the engine—deployed contracts, property tests, CLI shocks, published script hashes/tx IDs.
• M3 – 13,111 ADA: the vehicle—Android APK with 30+ tx across 10+ devices, crash/perf gates (P90 ≥20m crash-free, P95 and lt;60s confirmation).
• M4 – 8,706 ADA: the road test—two workshops, pilot dashboard, anonymized CSV, and a go/no-go memo so Catalyst only funds real rails if users actually succeed.

Clear, realistic targets (we pre-commit to these)

• 150–200 testnet wallets created
• 120–150 full loan cycles (lock → borrow → repay)
• 600–800 total contract calls
• 10+ liquidation demos in price-shock tests
• 0 invariant breaches; P95 confirmation All numbers are backed by script hashes and tx IDs on the testnet explorer.

Real users, right audience

East Africa is mobile-money first. We simulate mobile-money steps to validate the UX now (without expensive integrations), capturing SUS and task-success data that other Cardano teams can reuse.

Hard gates protect funds

Treasury is a 2-of-3 multisig (Nasser, Salm, Semir). No payout until public evidence is posted: hashes, addresses, tx IDs, CI/coverage, APK checksum, device-matrix results, pilot CSV. That makes waste hard and auditing easy.

Reviewer Concerns → Our Response

• “Why not integrate Airtel/MTN now?” Partner work dominates cost. First we prove users can complete flows and understand liquidations. If KPIs clear, we propose Phase-2 with data, not guesses.
• “Why pay for workshops?” Usability drives adoption. Two lean workshops produce CSV + SUS/task metrics and reusable copy/patterns—savings for every team that follows.
• “Why a Safety Lead?” Liquidation math and oracle checks are where DeFi loses money. A small spend on control policy + FMEA/HAZOP + property tests is far cheaper than one bad incident.

Bottom line

For 48,000 ADA, Cardano gets:

1. Audited lending contracts with invariants,
2. A price-shock simulator,
3. A production-ready Android testnet wallet, and
4. A public dataset + report proving what works for mobile money style users. It is evidence first, open, and reusable maximizing return on ADA while minimizing risk to Catalyst and the wider ecosystem.

[Required Acknowledgements] Consent & Confirmation

Terms and Conditions:

Yes