Numbers-first board document · content/business_review.md
QALS — Business Review for the Board
Prepared: 6 September 2026 • Prepared for: qalarc board / prospective investors Scope: Qalnet (private IOTA-derived chain), QALS credit (AU$1-backed B-QALS + growth G-QALS), Qal Compute marketplace, Qalx DeFi platform, and the staged bank/exchange path. Basis: qalarc internal research corpus (16 cloned IOTA-ecosystem repos, ~550MB; live web research dated 6 Sep 2026; 11 design documents). All figures from the corpus are cited to their source doc; all management estimates are marked [ASSUMPTION]; items the corpus could not verify are marked [UNVERIFIED].
1. Executive snapshot
What: QALS is qalarc's internal economy: a private, Apache-2.0 fork of the IOTA node (itself Sui-lineage: Move + EVM dual-VM, Starfish DAG-BFT consensus) run across qalarc's existing four-machine fleet, on which every qalarc app, AI agent, user and GPU settles in one currency. QALS exists in two classes: B-QALS, prepaid service credit born from and redeemable for AU$1 from a segregated full reserve, and G-QALS, a fixed-supply growth token supported by a revenue-funded floor (FloorVault). Around it sit a compute marketplace (Qal Compute) and a DeFi layer (Qalx).
Why now: (1) The infrastructure is free and mature — IOTA's 2025 "Rebased" rewrite plus the April 2026 Starfish consensus upgrade give us an enterprise-grade, audited chain we can fork today at zero licence cost. (2) IOTA's token collapse (price $0.041, rank #137, market cap $191M, all-time-low July 2026 — doc 01 §5) has made the strategic lesson cheap to learn: token price ≠ infrastructure value. We fork the node, not the tokenomics. (3) qalarc already owns the demand side — apps, agents, and GPU hardware that bill for AI/compute today through siloed Stripe-style billing. (4) Australia's regulatory path for the next steps (AUSTRAC VASP registration, closed-loop prepaid carve-outs) is documented and priced (doc 06).
Stage: Pre-build. Devnet achievable in weeks 1–4 at ~AU$0 cash cost (labour only); core contracts and fleet validators months 2–3 at ~AU$50–100/mo cloud (doc 00).
Capital needed: Bootstrap path AU$250k–400k over 18 months (legal, audit, AUSTRAC VASP program). Seed path AU$1.5M–3M to run Phase 2 (registered exchange surface, qAUD) at full speed. Details in §10.
The one-line thesis: a full-reserve prepaid credit system with breakage-free gift-card economics, wrapped in a private chain we control, monetising compute and AI services qalarc already sells — with no dependence on token speculation.
2. Company & asset context
qalarc operates a family of applications and infrastructure that today share no common ledger:
| Asset | What it is | Relevance to QALS |
|---|---|---|
| qalarc.com | AI services site | First credit-denominated billing surface |
| tradez.au | Trades/services marketplace | Invoice DataAnchors, escrowed settlement |
| doof.ing | Consumer/community property | Micro-payments, Doofs NFT drops |
| endispute.com.au | Dispute resolution | On-chain escrow + notarised evidence — natural fit (doc 04 §5) |
| Qalarc Hub (Signal/WhatsApp agent hub) | Agent orchestration on port 8769 | Every agent gets a DID + wallet + spend cap; first metering pilot |
| The Fleet (4 machines on Tailscale) | superlocal (96GB RAM, AMD 8060S 32GB), qalcachyminirig (GPU), bb-mini, cachyos-x8664 (spare) | Validators #1–4 + compute providers (doc 03 §5.3) |
Revenue base [ASSUMPTION]: qalarc is assumed to have an existing AI-agency/services revenue base that funds operations during the build. This review assumes that base covers founder time and the ~AU$50–100/mo infrastructure line but is not separately quantified in the corpus; the board should request the management accounts to pin this down.
What the corpus does establish: the fleet is real, in production use (a 30B-class LLM already served from the 8060S iGPU at ~38 tok/s per the local-GLM work), and residential electricity is ~AU$0.30/kWh — inference-sized jobs remain profitable against per-token API rates because typical API markup over raw compute is 5–20× (doc 07 §5).
Strategic asset: ten years of IOTA/Sui hardened open-source engineering (node, SDKs, Gas Station, identity framework, Stronghold) inherited under Apache-2.0, plus a public anchoring target (IOTA mainnet) for tamper-evidence (doc 03 §4, §5.5).
3. Market opportunity
Numbers we can source from the corpus, and where we must stay conservative:
| Market | Evidence (corpus) | Read |
|---|---|---|
| Australian BNPL/payments | Zip FY26: total income A$1,347.4m (+24.6%), TIV A$16.7b (+27.2%), cash EBTDA A$268.9m, 6.5m active customers, 97.4k merchants (zip.co/investors via doc 06 §1.1) | Proof that Australian consumers transact billions through credit/prepaid rails and that Zip itself is pivoting to "AI and agentic commerce" — the adjacency is real, not hypothetical |
| Stablecoins | Global stablecoin market ≈ US$250b (Jun 2025), +50% y/y, dominated by Tether US$162b / USDC US$61b (RBA FSR Oct 2025 via doc 06 §3.1) | The instrument class is growing fast; Australia is "a watcher, not a legislator (yet)" — regulatory clarity is the gate, not demand |
| Tokenised AUD | RBA: no retail CBDC ("no public interest case", Sep 2026); Project Acacia endorses private tokenised money; bank pilots (ANZ A$DC, NAB AUDN, Novatti AUDC) all pilot-scale [UNVERIFIED current status] (doc 06 §3.1) | No dominant AUD token exists. A closed-loop AU$-denominated credit is a legitimate early position, with partner-bank deposit tokens as the compliant endgame |
| AI inference / compute spend | Corpus contains no sourced AI-spend forecasts. Directional evidence only: API markup over raw compute 5–20× (doc 07 §5); Zip IR narrative shift to agentic checkout (doc 06 §1.1) | Conservative stance: we do not claim a TAM number we cannot source. Our entry is cost-side (sell our own GPU capacity at API-competitive prices), not TAM-side |
| DePIN GPU markets | Akash, Golem, Render, io.net cited as precedent with tokenomics lessons; no revenue/TVL figures verified in corpus (doc 07 §1, §5) | Treat as design precedent, not comparable financials. Our differentiation is being our own anchor tenant — no cold-start |
| Trade/commodity provenance | IOTA: TWIN live in Kenya customs; ADAPT with AfCFTA/WEF targeting 1.5B people by 2035; Salus tokenised tantalum on mainnet (doc 01 §4) | Validates object-ledger + provenance patterns at institutional scale; relevant to endispute and AI-output receipts, not a near-term revenue line for us |
Honest framing: the addressable opportunity for the first 18 months is qalarc's own billing plus a modest external compute/credit book in Australia — not global DeFi TAM. Every projection below is built bottom-up from throughput assumptions, not top-down from market size.
4. Product lines & revenue model
All services are paid for in QALS credit (B-QALS), priced in AUD via an on-chain PriceTable (doc 04 §4).
| # | Product line | Revenue mechanism | Rate (start) | Source |
|---|---|---|---|---|
| 1 | AI/compute services (inference, subscriptions, anchoring) | Service gross margin on credit consumed; B-QALS burned on use, AU$1 Reserve→Operations | 18–25% blended; 22% base [ASSUMPTION] | docs 03/04; 22% illustrative in doc 03-backing §6 |
| 2 | Qalx — parity rail (B-QALS ⇄ qAUD) | Fee on credit throughput | 1–5 bps | doc 05 §4 |
| 3 | Qalx — AMM (utility pool) | Treasury share of 0.25% fee (20 bps LP / 5 bps treasury) | 5 bps of volume | doc 05 §2.2/§4 |
| 4 | Qalx — redemption desk | Redemption fee (flows to FloorVault, backs G-QALS — not P&L) | 0.5% | doc 05 §4; doc backing §2 |
| 5 | Qalx — receivables vault (ComputeReceipt & invoice finance) | Flat fee on advances | 1.5–3% | doc 05 §2.3/§4 |
| 6 | Qalbook CLOB (Phase B) | Maker/taker | 0–2.5 bps, epoch-voted; trigger >AU$1M/mo internal flow | doc 05 §2.4/§4 |
| 7 | Qal Compute marketplace | Take on external provider settlements; burn-mint slice on all settlements | ~10% take [ASSUMPTION]; corpus does not fix a rate | doc 07 §2/§5 |
| 8 | Compute futures/vouchers | Capacity pre-sales at strike (internal first) | margin on capacity planning | doc 07 §3 |
| 9 | NFTs | ComputeReceipt NFTs (provenance, free to mint); Doofs PFP drops later | drop proceeds; episodic | doc 07 §2; doc 00 |
| 10 | Provider listing/verification | Fixed fees, external providers (Phase 3) | TBD | doc 05 §4 |
Structural point: revenue line 1 dominates for the first two years and is independent of QALS market price — it is prepaid credit consumed as services (doc 04 §7). Lines 2–6 are high-margin but small until throughput scales; line 4 deliberately funds the FloorVault rather than profit.
5. The AU$1 backing system as commercial moat
The corpus is unusually blunt here (backing doc §0): backing all 4.6B QALS at AU$1 would require AU$4.6B — impossible, and implying it without holding it is "a stablecoin lie (see every 2022 collapse)." The design instead makes the backing true where it is funded:
| Mechanic | Commercial meaning |
|---|---|
1 AUD in → 1 B-QALS minted; segregated client-money reserve; invariant AUD ≥ B-QALS × AU$1 enforced at mint (doc backing §2) |
Trust as a product feature: the chain is the liability ledger; monthly published proof-of-reserves. No bank-run insolvency risk — first redeemer is still fully served |
| Consumption burns B-QALS; AU$1 moves Reserve→Operations as revenue | Prepaid/gift-card economics without expiry gimmicks; revenue recognition is on-chain-auditable |
| Redemption at AU$1 − 0.5% | Cheap exit is what makes entry credible; the 0.5% funds the FloorVault |
| Rewards bought, never minted unbacked (marketing budget buys B-QALS at AU$1 from issuance) | Every reward is born backed — no hidden dilution of the credit class |
| FloorVault: 20% of service gross margin + 100% redemption fees + 50% of Qalx fees + 100% of slash takings, ratcheted floor, FloorBot bid at floor − ε (doc backing §3) | G-QALS has an honest, revenue-funded floor — published monthly ("coverage: 0.7% → target parity, funded by revenue, not promises") |
| Reserve tranche in AU T-bills once >AU$250k; yield accrues to FloorVault (doc backing §5) | Float income once the reserve is large — the classic prepaid economics, captured transparently |
Breakage economics — honestly stated: the design sets no expiry by default (breakage only as a regulatory provision), so we do not book gift-card-style breakage income. The commercial value of the float is (a) working capital in the interim, (b) T-bill carry to the FloorVault at scale, and (c) conversion: prepaid balances convert to services at 22% gross margin, better than the ~3.9% net transaction margin Zip earns on TIV (doc 06 §1.1).
Worked example from the corpus (backing doc §6): Month 1 — AU$40k top-ups, AU$28k consumed (revenue), reserve AU$12k vs 12k B-QALS outstanding; FloorVault +AU$3k (20% of margin) plus fees. Month 18 — AU$500k/mo top-ups at 22% margin → FloorVault +AU$22k/mo; with 60M G-QALS circulating the floor is AU$0.0004 — tiny but honest, and the doc says so openly.
Why this is a moat: the corpus's IOTA post-mortem (doc 01 §7) shows a decade of institutional wins (WEF, UK Cabinet Office, GLEIF) producing zero price support for an unbacked inflationary token. A full-reserve, redeemable, monthly-attested credit is the structural opposite — and it is the framing a sponsor bank, AUSTRAC, and "a grandmother" can all verify (doc backing §7).
6. Unit economics — worked scenarios
Assumptions (all [ASSUMPTION] unless sourced): 22% blended service gross margin (base; bear 18%, bull 25%); throughput = services consumed in AUD = credit throughput; parity rail fee 3 bps of throughput; AMM treasury share 5 bps on utility-pool volume (assumed 10% of throughput, live from month 4); external compute marketplace take 10% on external settled volume; receivables vault 2% average on advances outstanding; redemption rate 10% of top-ups (fee → FloorVault, excluded from P&L); NFT proceeds episodic and excluded from the table.
Credit throughput (AU$/month)
| Scenario | M1 | M6 | M18 | M36 | Anchor rationale |
|---|---|---|---|---|---|
| Bear | 20,000 | 60,000 | 150,000 | 250,000 | Internal apps only, slow external uptake |
| Base | 40,000 | 150,000 | 500,000 | 800,000 | M1 matches corpus worked example (AU$40k, doc backing §6); M18 matches corpus AU$500k/mo illustration |
| Bull | 60,000 | 300,000 | 900,000 | 2,000,000 | Compute marketplace hits traction + tradez/endispute books |
Monthly gross profit by line — BASE case (AU$/month)
| Line | M1 | M6 | M18 | M36 |
|---|---|---|---|---|
| Service margin (22%) | 8,800 | 33,000 | 110,000 | 176,000 |
| Parity rail (3 bps) | 12 | 45 | 150 | 240 |
| AMM treasury share | — | 8 | 25 | 40 |
| Compute take (10% external vol: 0/10k/100k/400k) | — | 1,000 | 10,000 | 40,000 |
| Receivables vault (2% on advances: –/–/50k/200k) | — | — | 1,000 | 4,000 |
| Total gross profit | 8,812 | 34,053 | 121,175 | 220,280 |
Monthly gross profit by scenario (AU$/month, service margin only + platform lines at base mix)
| Scenario | M1 | M6 | M18 | M36 |
|---|---|---|---|---|
| Bear (18% margin) | 3,600 | 10,800 | 27,000 | 45,000 |
| Base (22%) | 8,800 | 34,100 | 121,200 | 220,300 |
| Bull (25%) | 15,000 | 75,000 | 225,000 | 500,000 |
Cumulative 36-month base case
| Measure | Value |
|---|---|
| Cumulative credit throughput (ramp 40k→800k, ~linear) | ≈ AU$16.2M |
| Cumulative gross profit (22% + platform lines) | ≈ AU$3.6M |
| Implied breakeven vs lean cost base (~AU$45–60k/mo from month 4, §7) | ≈ month 20–24 |
| Reserve float at M36 (steady-state outstanding balance ~1 month of top-ups) | ≈ AU$800k → T-bill carry to FloorVault ≈ AU$30–40k/yr at ~4% [ASSUMPTION rate] |
Sensitivity: each ±AU$100k/mo of throughput at M36 = ±AU$262k/yr gross profit. The model is throughput-first; every other line is a rounding error until volume exists. The corpus's own numbers agree — parity-rail and AMM fees are bps-level by design (doc 05 §4).
7. Cost base
| Item | Bootstrap (AU$) | Funded/seed (AU$) | Source |
|---|---|---|---|
| Chain infrastructure (cloud VMs, RPC gateway, monitoring) | 50–100/mo | 300–1,000/mo | doc 00; doc 03 §5.3 |
| Engineering | founder time funded by agency base [ASSUMPTION] | 2–3 FTE ≈ 260k–390k/yr [ASSUMPTION salaries] | — |
External Move audit (qal_* packages, pre-revenue) |
US$30–80k one-off | same + QALS-denominated bug bounty | doc 05 §3.4 |
| Fintech legal opinions (Phase 1 prepaid framing; the "load-bearing wall") | 30–80k | 80–150k | doc 06 §2.4/§5 |
| AUSTRAC VASP program (AML/CTF build, consultants, registration) | 150k–400k | 150k–400k | doc 06 §5 Phase 2 |
| KYC vendor (FrankieOne-class) | per-verification, quoted A$1–4 at volume [UNVERIFIED pricing] | same | doc 06 §4.1 |
| Reserve attestation (quarterly AUP; annual audit at Phase 2+) | 20–40k/yr | 50–100k/yr | doc backing §2 [ASSUMPTION fees] |
| Contingency: Treasury digital-asset/stablecoin legislation passes | — | AFSL-variation project ~100k–300k, 6–12 months [UNVERIFIED estimate] | doc 06 §2.5/§5 |
Key property: the build cost of the chain itself is ~zero (Apache-2.0 fork, existing hardware, labour only). The cash cost base is dominated by compliance artefacts — audit and legal — which is exactly where a payments business should spend its first dollars.
8. Competitive landscape
| Competitor class | Who | Their position | Why QALS differs |
|---|---|---|---|
| Public IOTA EVM ecosystem | Pools (DEX), Swirl (liquid staking), Virtue (CDP), CyberPerp; LayerZero-connected (doc 01 §3) | Small but coherent DeFi on a $191M-cap chain; IOTA Foundation has pivoted wholly to trade infra | We use their code, not their market. No dependence on IOTA token demand; worst case we track MystenLabs/sui directly (doc 03 §7) |
| Public L1s (Sui-class) | Sui, Aptos et al. | Deep liquidity, big dev ecosystems | They sell speculation surface and shared security; we sell a controlled billing rail with gasless UX (Qal Pass) and zero public-trading risk pre-licence |
| DePIN compute | Akash, Golem, Render, io.net | Real markets, account-based chains, token-incentivised supply | They face cold-start and token-price-driven supply churn; we are our own anchor tenant with owned GPUs (doc 07 §5), Move-enforced escrow, and receipts with provenance |
| BNPL/consumer credit | Zip (A$16.7b TIV FY26); BNPL now LCCC/ACL-regulated since 10 Jun 2025 (doc 06 §1) | Licensed, scaled, moving into agentic commerce | We are prepaid, not credit, in Phase 1–2 — no ACL, no responsible-lending obligations until we choose Phase 3; first-party instalments sit outside the LCCC definition (doc 06 §1.2) |
| Australian exchanges/stablecoins | AUSTRAC-registered VASPs; Novatti AUDC (Stellar) [UNVERIFIED status]; bank pilots ANZ/NAB | Fragmented, pilot-scale AUD tokens; RBA direction is bank-issued deposit tokens | Our qAUD is closed-loop until VASP registration; the partner-bank deposit-token endgame (Acacia-endorsed) is explicitly on our roadmap rather than fought (doc 06 §3.2 Option B) |
| Traditional billing (Stripe, SaaS invoicing) | Incumbent default | Ubiquitous, trusted | Cannot express machine-to-machine metered payments, agent spend caps, or on-chain escrow — the exact gap QALS fills (doc 04 §1) |
Positioning sentence: the only competitor that does full-reserve AU$ prepaid credit on a private object-ledger with escrowed compute settlement is nobody. Each incumbent owns one slice; the integrated stack is the differentiation.
9. Risk register (top 10)
| # | Risk | Likelihood | Impact | Mitigation |
|---|---|---|---|---|
| 1 | Regulatory reclassification of QALS as a financial product (stablecoin/security/PPF boundary moves) | Medium–High | Severe | Phase 1 strictly closed-loop prepaid; written fintech opinion before any external sale/transferability; AUSTRAC VASP before exchange surfaces; watch Treasury digital-asset bill (doc 06 §2.5) |
| 2 | Demand shortfall — throughput never reaches AU$500k/mo | Medium | High | Anchor-tenant model: qalarc's own apps and hub metering are the first book; costs are near-zero so the downside is cheap to carry |
| 3 | Reserve/custody failure or loss of confidence | Low | Severe | Full reserve by construction (mint gate asserts invariant on-chain); segregated client-money account; monthly published reconciliation; >0.5% discrepancy auto-pauses mint (doc backing §5) |
| 4 | Oracle committee compromise (3-of-5 reserve signers) | Low | High | Signers on separate machines/owners + one external key; 2 can trigger investigate mode; rate limits (AU$250k/day mint cap Phase 1) cap blast radius (doc backing §5; doc 05 §3.1) |
| 5 | Fork maintenance burden as upstream IOTA moves fast | Medium | Medium | Changes confined to genesis + qal_* Move packages; monthly upstream tracking; Sui fallback is portable (doc 03 §7) |
| 6 | Validator outage (home fleet on residential links) | Medium | Medium | Starfish is designed for lagging validators (that's why we chose it); cloud VM in committee from day one; Tailscale mesh already proven (doc 03 §5.3) |
| 7 | Key management / agent key compromise | Medium | High | Stronghold secure storage; per-agent epoch spend caps enforced by chain (not prompts); Hierarchies revocation ~2s kill switch; revocation drills (doc 03 §7; doc 04 §3) |
| 8 | Security bug in inherited code | Medium | High | We inherit Sui/IOTA audits; external Move audit of our packages pre-revenue (US$30–80k); pen-test before real value moves; bug bounty in QALS (doc 05 §3.4) |
| 9 | Skill ramp (Move/Rust) | Medium | Medium | Sui-class docs and workshops; EVM fallback surface exists; start with small packages (doc 03 §7) |
| 10 | Key-person / concentration risk (founder-led, 4 machines) | Medium | High | Documented runbooks; external Oracle Committee signer; seed path adds team depth; this review itself is a governance artefact |
Standing lesson encoded in the register (doc 01 §5–§7): IOTA hit all-time-low ($0.0311, 31 Jul 2026) in the same quarter its technology shipped Starfish and its partnerships hit WEF/UK-government grade. Token price is not an infrastructure KPI. QALS therefore carries no revenue line that depends on token appreciation — G-QALS upside is a treasury/alignment instrument with a revenue-funded floor, not a business model.
10. Milestones & funding options
Milestones (doc 00; doc 03 §6; doc 05 §5)
| Window | Milestone | Cash cost |
|---|---|---|
| Weeks 1–4 | qalnet-dev-1 devnet on superlocal; fork renamed; QALS genesis; first DataAnchor from agent hub |
~AU$0 (labour) |
| Months 2–3 | The Fleet validators + qal_credit/qal_compute/qal_data core contracts; hub LLM metering live |
AU$50–100/mo |
| Months 4–6 | All apps integrated; compute settlement + ComputeReceipt NFTs; Qalx v0 (redemption desk + proof-of-reserves) → v1 (AMM + FloorBot) | + audit US$30–80k |
| Months 6–9 | Phase 1 legal opinion closed; AUSTRAC enrolment → VASP program build | A$150k–400k program |
| Months 9–18 | Qalx Phase 2 surfaces (post-VASP); qAUD partner rails; Doofs NFT drop; Qalbook trigger check (>AU$1M/mo flow) | per §7 funded column |
| Months 18–36 | Receivables vault at scale; external compute providers; Phase 3 credit decision (ACL) | A$300k–1M+ if ACL pursued (doc 06 §5) |
Funding options
| Path | Raise | What it unlocks | What it forgoes |
|---|---|---|---|
| Bootstrap (agency-funded) | AU$250k–400k over 18 months [ASSUMPTION availability] | Phase 1 + audit + VASP registration at a deliberate pace; zero equity dilution; G-QALS investor bucket (10%, doc 03 §5.2) retained | Slower: CLOB deferred, external compute providers deferred, single-threaded team |
| Seed | AU$1.5M–3M [ASSUMPTION valuation not modelled] | 2–3 engineers, full Phase 2 in 9 months, Qalbook port + audit, external provider acquisition, marketing budget that doubles as FloorVault funding | Dilution; investor pressure toward earlier public surfaces — must be governed by the compliance gates, not the cap table |
Recommendation: bootstrap to the month-4 milestone (working devnet + metering + audit quote), then decide the seed question with real throughput data. The corpus's cost structure makes this unusually cheap to stage.
11. Decision asks for the board
- Approve Phase 1 build (weeks 1–4 devnet → month 6 core contracts) with a capped Phase 1 budget of AU$50k cash (legal review + audit deposit), all other costs from existing operations.
- Approve the token framework: fixed 4.6B QALS supply, no protocol inflation, and the 8-bucket allocation incl. 20% treasury / 15% team (4y, 1y cliff) / 10% investors (doc 03 §5.2) — with the standing rule that no public sale or listing occurs before the licensing gates clear.
- Ratify the full-reserve covenant: B-QALS minted only against cleared AUD; monthly published proof-of-reserves; FloorVault ratchet that can rise but never fall; redemption fee 0.5% to FloorVault — as published policy, not engineering discretion.
- Choose the funding path at the month-4 gate: bootstrap (AU$250k–400k) vs seed (AU$1.5M–3M), decided against measured internal credit throughput.
- Appoint the external Oracle Committee signer and engage Australian fintech counsel for the written Phase 1 opinion (the eight questions in doc 06 §5 are the brief) — before any QALS is sold to anyone outside qalarc.
12. Disclaimer & sources
NOT LEGAL ADVICE. This review is a commercial synthesis of internal research documents that were themselves prepared from public web sources by an automated agent. Australian financial services law is complex, fact-dependent and changing; nothing here should be relied on without written advice from an Australian financial-services lawyer and, where relevant, ASIC, APRA, AUSTRAC and ATO input. Items the underlying research could not verify against a primary source are marked [UNVERIFIED]; management estimates are marked [ASSUMPTION].
Primary corpus (all dated 2026-09-06): 00_EXECUTIVE_SUMMARY.md • 01_iota_deep_research/STATE_OF_IOTA_2026.md (IOTA price/market data via CoinMarketCap; IOTA Foundation blog posts; Wikipedia) • 03_qals_architecture/QALS_ARCHITECTURE.md • 03_qals_architecture/QALS_BACKING_DESIGN.md • 04_credit_token_platform/CREDIT_PLATFORM.md • 05_defi_trading/QALX_DEFI_PLATFORM.md • 06_bank_exchange/BANK_EXCHANGE_PLAN.md (Zip IR, APRA, AUSTRAC, RBA/Acacia sources listed in that document) • 07_compute_marketplace/COMPUTE_MARKETPLACE.md. External claims (Zip FY26 metrics; stablecoin market size; regulatory dates) inherit the citation status of those documents.
Prepared by qalcode document-author agent, 6 September 2026. ~4,100 words.