QALS Wiki · the qalarc networkgenerated 2026-09-06 · qalcode autonomous research

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

  1. 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.
  2. 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.
  3. 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.
  4. 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.
  5. 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.md01_iota_deep_research/STATE_OF_IOTA_2026.md (IOTA price/market data via CoinMarketCap; IOTA Foundation blog posts; Wikipedia) • 03_qals_architecture/QALS_ARCHITECTURE.md03_qals_architecture/QALS_BACKING_DESIGN.md04_credit_token_platform/CREDIT_PLATFORM.md05_defi_trading/QALX_DEFI_PLATFORM.md06_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.