# Ouda Developments: a whitepaper on documenting maintenance deposits with blockchain

> How Ouda documents maintenance deposits on blockchain: money stays in an Egyptian bank, owners verify every pound.

Source: https://ouda-developments.com/en/ouda-maintenance-blockchain-whitepaper/
Language: en
Publisher: Ouda Developments — https://ouda-developments.com

---

[Press Release](https://ouda-developments.com/en/media-coverage/)

Ouda DevelopmentsWhitepaper · Version 1.018 November 2026Design stage

## **Every maintenance pound** backed by mathematical proof

How Ouda Developments uses blockchain technology to document maintenance deposits and expenses, so every owner can verify them personally without having to trust anyone. The money stays in an Egyptian bank, in Egyptian pounds.

**SHA-256** a digital fingerprint for every entry

**32 bytes** daily seal on a public network

**2256** possibilities facing a forger

**0 EGP** allowed gap vs. the bank statement

## ★The idea in one minute

Owners pay a large maintenance deposit, then have no independent way to see where it went. Ouda’s new system changes this: every pound entering or leaving the maintenance account becomes a sealed digital entry that cannot be edited, deleted or back-dated without being exposed immediately.

🏦

#### Money stays in the bank

The deposit is held by a licensed Egyptian bank in Egyptian pounds. No crypto, no owner wallets.

🔏

#### Fingerprint on the network

Once a day, one fingerprint of the whole ledger is sealed on a public blockchain that Ouda does not control.

📱

#### Owners verify for themselves

With one tap, the owner’s phone recomputes the fingerprint and compares it with the sealed one: “Match” or “No match”.

1

**Expense or deposit** invoice, amount, category, approvals

2

**Fingerprint** SHA-256 of the entry

3

**Chain** each entry carries the previous one

4

**Merkle tree** the day’s entries in one root

5

**Public daily seal** 32 bytes on a public network

6

**Owner check** from the phone, no middleman

## 1The problem: trusting the maintenance deposit

The maintenance deposit is a percentage of the unit price that every owner pays, and it is managed for many years after handover. Today the only source of trust is the word of whoever manages it.

To show the scale, here is a hypothetical example with simple numbers (not the figures of any actual project):

| Item (hypothetical example) | Value |
| --- | --- |
| Unit price | EGP 1,000,000 |
| Maintenance deposit | 8% = EGP 80,000 per unit |
| Number of units | 100 units |
| Total deposit | EGP 8,000,000 |
| Operation and maintenance | Bravo Facility Management |

🙈

#### Can’t see

The books stay inside the company; at best an annual summary arrives.

❓

#### Can’t verify

Even when shown the ledger, the owner can’t tell whether it is last month’s ledger or an edited one.

🧾

#### Audits are a snapshot

An auditor sees what is presented on audit day, not what happened all year.

**The idea borrowed from Bitcoin:** replacing “trust the intermediary” with “verify it yourself with mathematics”. We take only that part and leave the currency and mining out entirely.

## 2The trust model: money in the bank, proof on the network

The system separates where the money is from where the evidence is. The money never moves on the blockchain, and the evidence never stays in the company’s hands alone.

| Party | What it holds | Role |
| --- | --- | --- |
| Licensed Egyptian bank | The deposit in EGP, a separate account per project | The bank statement is the final reference for the balance |
| Ouda ledger | Every movement with its documents and approvals | Records each event as it happens |
| Public network | One fingerprint per day (32 bytes) | A dated witness Ouda does not control |
| Owners and auditors | The verification tool | Compare the ledger with the seal and the bank statement |

To hide any tampering, someone would have to defeat **at the same time** the mathematics, a public network they do not own, and the bank statement.

## 3The entry: the system’s building block

Every financial movement is recorded as a single “entry” with fixed fields. There are four entry types: deposit, investment return, maintenance expense and correction.

| Field | Content | Example |
| --- | --- | --- |
| Sequence number n | Increases by one per entry, no gaps | 000412 |
| Project | Project code | CANARY |
| Type | Deposit / return / expense / correction | Expense |
| Amount | In piasters (whole number) | 4,850,000 piasters = EGP 48,500 |
| Category | Code from a fixed chart of categories | Elevators |
| Supplier | Code from the supplier register | V-017 |
| Document fingerprint | H(invoice PDF) | 9f2c… |
| Bank reference | Transaction number on the statement | TRX-88213 |
| Approvals | Digital signatures (section 7) | 2 of 3 |
| Previous fingerprint | h(n−1) | 41ab… |

**Why piasters?** Decimal numbers in computers create tiny rounding differences, enough to break reconciliation with the bank. **And nothing is ever deleted:** a mistake is fixed by a new correction entry that states the reason, so owners see both the error and its fix.

## 4The digital fingerprint (hash)

A mathematical function that turns any text or file, whatever its size, into a fixed-length 256-bit number. We use SHA-256, the same function Bitcoin is built on, and a published standard (FIPS 180-4).

Equation 1 · Fingerprint

h=H(r)∈{0,1}256

In plain words: Any entry r goes into the function H and a fixed-length number h comes out. The same entry always gives the same fingerprint, on any computer in the world.

A real example computed with SHA-256. The two entries differ by **one piaster** , and the fingerprints are completely different:

CANARY|000412|EXPENSE|485000 *0* |ELEVATORS
*→ 197c1e08750ff974875cd1aeb55633553b69bf630945b4a21fb1e4fda65aa0f4*

CANARY|000412|EXPENSE|485000 *1* |ELEVATORS
*→ 8e50545f6a87d83ef3e34eccea48b930f8f3cf747c310d544dc244378c478123*

🎯

#### Deterministic

The same input always gives the same output.

⚡

#### Highly sensitive

Changing a single piaster changes almost the whole fingerprint.

🔒

#### One-way

The entry can’t be recovered from its fingerprint, and no other entry can be forged with the same one.

## 5The chain of entries: nothing deleted, nothing slipped in

Each entry carries the fingerprint of the one before it, so the entries link together like the links of a chain.

Equation 2 · Linking

hn=H(hn−1‖n‖rn)

In plain words: Each entry’s fingerprint is computed from its content, its number and the previous entry’s fingerprint. Change an old entry and every fingerprint after it changes.

**Entry 410** `h=7c21…`

**Entry 411** carries 7c21…`h=41ab…`

**Entry 412** carries 41ab…`h=197c…`

**Entry 413** carries 197c…`h=e05d…`

| Tampering attempt | Result |
| --- | --- |
| Editing an old entry | Exposed its fingerprint and every later one change |
| Deleting an entry | Exposed a gap in the sequence numbers and a broken chain |
| Inserting a back-dated entry | Exposed there is no slot for it without recomputing everything after |

Whoever holds the ledger could in theory recompute the whole chain. That is why we need an **outside witness** that keeps the fingerprint on its day, and that is the role of the public network.

## 6Merkle tree and the daily seal on a public network

At the end of each day, the fingerprints of that day’s entries are combined in a “Merkle tree”, a method invented by Ralph Merkle and used by Bitcoin itself.

Equation 3 · Tree

mparent=H(mleft‖mright)

In plain words: Every two fingerprints are merged into one, and so on upwards, until a single number remains: the “root” R(d), which stands for all of the day’s entries.

Equation 4 · Daily seal

Ad=H(Ad−1‖Rd‖d‖project)

In plain words: Today’s root is linked to yesterday’s seal, and the result A(d), just 32 bytes, is written to a public network. From then on, that day’s ledger has a witness nobody can erase.

**Why a tree and not a list?** Because an owner can prove their entry is inside the day with a very short “path”, without downloading the other entries:

Equation 5 · Proof size

|π|=⌈log2N⌉×32 bytes

In plain words: The proof grows very slowly: even with a million entries it is only 640 bytes, smaller than a short text message.

| Entries | Fingerprints in the path | Proof size |
| --- | --- | --- |
| 1,000 | 10 | 320 bytes |
| 10,000 | 14 | 448 bytes |
| 1,000,000 | 20 | 640 bytes |

## 7Approvals: no expense on one person’s signature

Every expense needs approvals from more than one party, and each approval is a digital signature with its owner’s private key (Ed25519, RFC 8032).

Equation 6 · Signature

σi=Signski(H(rn))

In plain words: Each signer signs the entry’s fingerprint with a secret key. The public key is published, so any owner can confirm the signature is genuine.

Equation 7 · Acceptance rule

valid(rn)⟺∑i=1m1[Verifypki(H(rn),σi)]≥k(amount)

In plain words: An entry is accepted only if the number of valid signatures reaches the required threshold, and the threshold rises with the amount.

| Expense amount | Approvals | Signers |
| --- | --- | --- |
| Below a small threshold (to be set) | 2 of 3 | Bravo operations manager, Ouda finance, owners’ representative |
| Above the small threshold | 3 of 3 | All three |
| Annual or exceptional contracts | 3 of 3 + prior notice | All three |

**The fixed principle:** no single party can spend on its own, not even Ouda. Thresholds and signers are set with the owners’ association.

## 8Balance equations and bank reconciliation

The ledger computes the balance at any moment from the entries alone, then it is compared with the bank statement every month.

Equation 8 · Balance

Bt=B0+∑n≤tDn+∑n≤tIn−∑n≤tEn+∑n≤tCn

In plain words: Balance = opening + owner deposits (D) + investment returns (I) − expenses (E) ± corrections (C).

Equation 9 · Reconciliation

Δmonth=Bmonthledger−Bmonthbank=0

In plain words: At the end of each month, the ledger balance minus the bank balance must equal zero, to the piaster. The result itself is recorded and sealed, and any gap is shown to owners with its reason.

Equation 10 · Principal protection (proposed)

ρyear=∑Eyear∑Iyear≤1

In plain words: The goal is for the deposit’s returns to cover the year’s expenses. If the ratio goes above 1, the principal is being spent, and owners will see it for themselves.

**Illustrative example:** in the hypothetical example (100 units, total deposit EGP 8 million), assuming an annual return r for illustration only:

| Assumed return r | Annual return | Per unit per month |
| --- | --- | --- |
| 10% | EGP 800,000 | about EGP 667 |
| 15% | EGP 1,200,000 | EGP 1,000 |

Both rates are assumptions to explain the equation, not a forecast of interest rates.

## 9How an owner verifies

Owners don’t need to understand the equations; they work for them in the background. From the community app or the owners’ portal:

1

**Open the expense** amount, category, supplier, invoice, approvals

2

**Tap “Verify”** the phone computes the fingerprint itself

3

**Rebuild the root** from the Merkle path (10 to 20 fingerprints)

4

**Read the seal** directly from the public network

✓

**“Match”** or “No match”

Equation 11 · Verification

x0=h,xj=H(xj−1‖sj),accept⟺xL=Rd

In plain words: The phone starts from the entry’s fingerprint and merges it step by step with the neighbouring fingerprints until it reaches the root. If it matches the sealed root, the entry is 100% intact.

**The verification tool is open source:** any programmer, auditor or journalist can run it independently of Ouda’s app. That is the difference between a “transparency report” and “transparency proof”. An independent monitor also recomputes the whole chain every day.

## 10Privacy

No personal data ever reaches the public network. What is published there is one irreversible fingerprint per day.

🎭

#### Pseudonyms

Owners appear as codes, not names or unit numbers. Each owner sees their own deposits; everyone sees expenses.

🧂

#### Salted fingerprints

Sensitive data is sealed together with a secret random number (salt), so it can’t be guessed.

📄

#### Invoices stay with us

The file stays on Ouda’s servers and is shown to owners; only its fingerprint enters the chain.

Equation 12 · Salting

c=H(salt‖owner_id)

In plain words: The owner’s identity is mixed with a secret random number before fingerprinting, so nobody can trace it back to a name or a unit.

The design follows the data-minimisation principle of Egypt’s Personal Data Protection Law No. 151 of 2020 and will be reviewed by counsel as part of the legal opinion.

## 11Security math and the system’s limits

To change an entry while the sealed fingerprint stays valid, a forger must find a different entry with the very same fingerprint.

Equation 13 · Chance of success

p=2−256≈8.6×10−78

In plain words: The chance that a single attempt succeeds is less than one in a number with 78 zeros.

Equation 14 · Time required

T≈22561021≈1.16×1056 s≈3.7×1048 years

In plain words: Even with computing power on the scale of the entire Bitcoin network (10²¹ attempts per second), a forger would need unimaginably longer than the age of the universe (about 1.4 × 10¹⁰ years).

And because the fingerprint is sealed on a public network, detecting tampering is not a matter of probability: **the first person to recompute it sees it with 100% certainty.**

**What the system protects against, and what it doesn’t.** Overpromising about technology destroys the very trust we want to build, so here are the limits plainly:

| Risk | Does the system stop it? | How |
| --- | --- | --- |
| Editing an amount after recording | Fully prevented | The sealed fingerprint no longer matches |
| Deleting an expense or deposit | Fully prevented | Gap in the sequence and a broken chain |
| Back-dated entry | Fully prevented | The daily seal on the public network |
| Expense approved by one person | Prevented | Minimum signature threshold |
| Gap between ledger and bank | Exposed monthly | Published reconciliation rule |
| Inflated invoice from the start | Exposed to owners | Visible invoice, external auditor, price comparison |
| Collusion of all signers | Not prevented | An owners’ representative among signers reduces the risk |

**Bottom line:** blockchain guarantees that what was recorded can’t change. Whether what is recorded is right in the first place is guaranteed by multiple approvals, owner oversight and the external auditor.

## 12Cost

The network cost is negligible, because we write only one transaction per project per day, however many entries there are.

Equation 15 · Network cost

Costyear=Np×365×f

In plain words: Projects × days in a year × fee per transaction. Four projects = 1,460 transactions a year, so even at a full dollar per fee the cost is USD 1,460 a year for the whole group.

The real cost is in software, operations and audit, to be set with the technology partner. Network fees are paid by an external technology partner that invoices Ouda in Egyptian pounds, so neither Ouda nor the owners hold any cryptocurrency. This structure is subject to counsel’s approval.

## 13Governance and legal compliance

The system will not go live before a written legal opinion from an Egyptian law firm specialised in banking and fintech.

1. Does recording fingerprints only on a public network fall outside crypto-asset trading under Central Bank Law No. 194 of 2020?
2. Is paying network fees through a technology partner invoicing in EGP sound?
3. What wording is allowed in announcements, and may the network be named?
4. Alignment with Personal Data Protection Law No. 151 of 2020.
5. Is prior approval from any regulator required?
6. Clauses needed in the buyer contract and the owners’ association bylaws.

| Party | Role |
| --- | --- |
| Ouda Developments | System owner, funds it, member of the approvals |
| Bravo Facility Management | Operator: records expenses and signs them |
| Owners’ association / representative | Independent signer, approves the annual budget |
| Egyptian bank | Holds the funds; its statement is the final reference |
| External auditor | Annual audit; its report is sealed in the ledger |
| Technology partner (to be selected) | Builds the system and pays network fees |

Even the system’s own rules, such as approval thresholds or the list of signers, are changed only through a sealed entry. **Nothing changes in the dark.**

## 14Implementation plan

The system is at the design stage today and will not record real money before the second gate: a written legal opinion and an independent security review.

#### ① Design — now

- Whitepaper 1.0
- Chart of categories
- Approval thresholds

◆ Gate: design approved and technology partner selected

#### ② Prototype

- Test data only
- Open verification tool
- Trial with Bravo

◆ Gate: written legal opinion and independent security review

#### ③ First operation

- First project handed over
- Real money
- Owners’ app

◆ Gate: zero bank gap for 6 consecutive months

#### ④ Rollout

- All group projects
- Annual external audit
- Public report to owners

Every public announcement about the system is tied to a stage actually delivered, not a promise. Gate criteria are proposed and will be agreed with partners.

## ?Owners’ FAQ

Will my money be turned into cryptocurrency?

No. The deposit stays in an Egyptian bank in Egyptian pounds; the public network only receives a digital fingerprint.

Do I need a wallet or to understand blockchain?

No. Verification is one tap in the owners’ app.

Will my name or unit number be visible online?

No. No personal data leaves the ledger, and owners appear as pseudonymous codes.

If an entry is wrong, is it deleted?

No. A new correction entry is added that everyone can see, and the original entry stays.

Is the system live now?

It is at the design stage and will not record real money before the legal opinion and the independent security review.

### What is visible is hard to cheat

The system doesn’t make fraud impossible, but it makes it visible, permanent and exposed. A developer’s responsibility doesn’t end at key handover, and this system makes it measurable.

[Discover Canary](https://ouda-developments.com/en/canary-compound/)

[Call 17793](tel:17793)

### References

1. S. Nakamoto, “Bitcoin: A Peer-to-Peer Electronic Cash System”, 2008.
2. S. Haber, W. S. Stornetta, “How to Time-Stamp a Digital Document”, Journal of Cryptology, 1991.
3. R. C. Merkle, “A Digital Signature Based on a Conventional Encryption Function”, CRYPTO 1987.
4. NIST, FIPS 180-4: Secure Hash Standard (SHA-256), 2015.
5. IETF, RFC 8032: Edwards-Curve Digital Signature Algorithm (Ed25519), 2017.
6. IETF, RFC 6962: Certificate Transparency, 2013.

Legal references: Egyptian Central Bank and Banking Sector Law No. 194 of 2020; Personal Data Protection Law No. 151 of 2020. Financial figures in this paper are illustrative examples, not the figures of any actual project.

Ouda Developments was founded in 2019 and builds on Ebdaa’s track record since 2001: 1,065 units delivered in 130 buildings on 555,079.65 m². Hotline 17793 · info@ouda-developments.com · a member of [Khaled Ouda Group](https://khaledoudagroup.com/). [اقرأ النسخة العربية](https://ouda-developments.com/ouda-blockchain-maintenance-whitepaper/).


<!-- Page uncached by LiteSpeed Cache 7.9.3 on 2026-11-18 10:12:07 -->