Amazon Clone Development Company: Who Should Build Your Marketplace
This page describes proposed scope and due-diligence criteria for a build in this category, agreed in writing per project. Nothing here is a finished product, and no client engagement for this platform is claimed.
Bring the checklist below. We would rather answer it before you buy than after.
Agency vs Freelancer vs the Proposed Omnistu Build
Three ways to get a multi-vendor marketplace, and what each one really asks of you. The middle column is proposed scope, agreed in writing per project - not an implemented product. No prices or timelines are published for any route; each is quoted in writing.
| Criteria | Freelancer / generic script | Omnistu proposed build | Custom dev agency |
|---|---|---|---|
| Time to launch | Unknown / DIY | A proposed deployment window, agreed in writing per project | Ask for a quoted milestone plan in writing before comparing routes |
| Multi-vendor | Ask whether vendor identity, approval and a vendor panel are native or a plugin | Vendor identity with KYC, approval, own panel and app | Ask for the multi-vendor scope in writing, with what is excluded |
| Delivery fleet | Ask whether delivery exists at all, or is out of scope | Agent identity, assignment feed, live location, OTP handover | Ask whether the fleet is in scope or quoted as a second product |
| Cash on delivery | Ask how cash is recorded, whose balance it sits on, and how it reconciles | Per-agent balance, dedicated transaction table, operator settlement | Ask how cash reconciliation is scoped and tested |
| Money model | Ask how many wallets exist and whether every movement writes a history row | Four wallets, append-only history per movement | Ask who on the team has built wallet ledgers before |
| Refund correctness | Ask what a full refund reverses: payment, commission, loyalty | Own status machine, commission reversed, loyalty reversed | Ask for the refund reversal behaviour in writing |
| Tenant isolation | Ask where tenant isolation is enforced | Repository-level scoping keyed on vendor and agent identity | Ask whether isolation is at repository level or per-query |
| Source code ownership | Ask whether any file is encrypted, obfuscated or licence-checked | Full backend and all three Flutter projects, no licence callback | Typically yes - confirm exactly what transfers |
| Cost | Low upfront cost - price the rows above before signing | A fixed quote for agreed scope, in writing | Scoped and quoted per engagement |
The proposed stack is deliberately conventional - PHP, Laravel, MySQL, Redis and Flutter - because the developers who maintain it after handover should already know it. That is a procurement consideration as much as a technical one.
Questions Worth Asking Any Provider
Ask these before you sign. The answers separate a marketplace from a storefront.
Show me a refund, fully
Not the customer credit - the whole reversal. Does commission come back off your books? Do loyalty points reverse? If a refund only moves money one way, you inherit a reconciliation problem.
Where does cash go?
An agent collects notes at a door. Ask which table that lands in, whose balance it belongs to until settlement, and how it reconciles to the operator wallet. "It is marked as paid" is not an answer.
How many wallets are there?
Customer, vendor, agent and operator are four different liabilities. One balance column means you cannot say who is owed what without a spreadsheet.
Can vendor A see vendor B?
Vendors are tenants. Ask where isolation is enforced - repository level, or a WHERE clause someone has to remember to add on every new query.
What stops a double order?
A second tap on a slow connection should not create a second order or reserve stock twice. Ask about the idempotency check, not about response times.
Which certifications do you hold?
Most imply more than they hold. We hold neither ISO 27001 nor SOC 2 and claim neither; the proposed build is designed to the controls and ready for the evidence gathering. Ask others to be equally specific.
The Development Process, in Order
What we do, in the order we do it. The sequence and its exit criteria are agreed in writing per project; no fixed day counts are published.
Step 1 - Scope, market and supply check
We confirm your market, your gateways, whether cash on delivery matters, and where your first vendors come from. A marketplace with no supply is a catalog of nothing, and we would rather say so before you pay.
Step 2 - Brand across six surfaces
Name, logo, splash screen and colour scheme on the storefront, vendor panel, operator console and all three Android apps, plus commission bands, zones, zip allowlists and tax classes.
Step 3 - Deploy and integrate
Deployment to your infrastructure, then your gateway credentials, SMS and email provider, Firebase project and any AI provider key connected and tested.
Step 4 - Run a full order, including cash
An order taken through split, dispatch, OTP handover and settlement - and a cash order run end to end - so you watch the money reconcile before a customer ever does.
Step 5 - Console and finance handover
The operator console walked through with the people who will run it: vendor approval, moderation, commission bands, coupons, payout processing and the wallet ledgers.
Step 6 - Source transfer and support
The backend and all three Flutter projects transfer to your repository, with reversible migrations and no licence callback. Support continues after go-live.
Red Flags That Mean Walk Away
We would rather you use this list on us than skip it.
Cash on delivery treated as an ordinary payment method
Cash collected at a door is a liability carried by a person, not a payment event. If it lands in the same table as card payments, the gap surfaces at year end.
One balance column doing the work of four wallets
Customer refund, vendor payout, agent cash and your commission are four different liabilities. One column cannot answer who is owed what.
A refund that keeps your commission
If the reversal returns the customer's money but leaves the commission on your books, every refund quietly overpays someone and vendors compare notes.
Tenant isolation by remembered WHERE clause
If isolation depends on every future query remembering to filter, one missed clause exposes one vendor's orders to another.
Delivery presented as a map screen
A map is not a fleet. Ask whether a delivery agent is an identity with KYC, a wallet and an approval state, or a flag on a customer account.
A certificate claim they cannot show you
An ISO 27001 or SOC 2 claim without the certificate is a mood, not a control. We hold neither and claim neither.
Client logos they will not let you verify
Unverifiable case studies are borrowed credibility. Ask for a reference you can actually call.
Source code encrypted or phoning home
Obfuscated files or a licence callback mean you are renting the platform you bought. Ask whether the matching, settlement and ledger logic transfers unobfuscated.
What a Multi-Vendor Marketplace Has to Get Right
Six things that are invisible in a demo and expensive at year end. Each is proposed design, agreed in writing per project.
Ledger integrity
Append-only histories written inside the same transaction as the balance change, so any balance can be re-derived. Without it, a disputed figure is unanswerable.
Cash reconciliation
Cash written to the collecting agent's balance and settled to the operator wallet, in its own transaction table. Without it, doorstep money and bank money never agree.
Tenant isolation
Repository-level scoping keyed on vendor and agent identity. Without it, one missed clause exposes one vendor's orders to another.
Refund symmetry
A refund status machine that prevents double-spend and reverses commission and loyalty. Without it, every refund quietly overpays someone.
Payment integrity
Gateway-hosted payment fields so no card number reaches the server, with webhook signature verification. Without it, an endpoint marks orders paid for anyone who finds it.
Delivery accountability
Status checkpoints and a one-time code verified at the door. Without it, "delivered" is a claim rather than a record.
Four Security Problems, Not One
Security on a marketplace is not a single control. Money moves between four wallets, vendors are tenants who must not see each other, customers hand over addresses and payment intent, and delivery agents carry cash and a live location trail that is itself sensitive. Every control below is proposed scope, not implemented or verified by this preview.
Money - ledgers that cannot be silently edited
Append-only ledgers are written inside the same transaction as the balance change, so a movement and its record cannot diverge. Any balance across the four wallets can be re-derived from history.
Tenancy - vendors cannot see each other
Isolation is enforced at repository level, keyed on the vendor and agent identity, rather than depending on every future query remembering to filter.
Payments - no card number reaches the server
Payment fields are gateway-hosted and webhooks are signature-verified, which keeps the heaviest part of payment compliance off your infrastructure entirely.
Agents - location and contact data segregated
A delivery agent's live location trail and emergency contact details are sensitive in their own right, and are visible to the agent and the operator alone.
Proposed controls, named individually: Argon2id password hashing, CSRF protection, parameterized queries, HTML sanitization, webhook signature verification and append-only audit ledgers with actor and timestamp. No certificate is held or claimed: the proposed build is designed for a penetration test and for the evidence gathering ISO 27001 or SOC 2 require. Several items are deployment-layer responsibilities we configure but you own after handover, and they are named as such rather than folded into a compliance claim.
Modelled Reference Configurations
Two illustrative configurations for very different commerce markets, drawn only from capabilities in the proposed base scope.
Read this before the models below
These are not client engagements: we have not yet deployed this multi-vendor commerce platform for a named client, and nothing here is presented as one. They are configurations, not outcomes - no revenue, traffic or growth figures are claimed, because none have been measured. And nothing in either model depends on custom work beyond the proposed base scope.
India - COD-led multi-vendor marketplace
What this market makes hard: cash closing the order at the door days after it was placed; tax rates differing by category and state; buyers who do not shop in English; delivery reaching some pin codes reliably and others not at all.
What an operator would want: take cash without inventing the accounting for it in a spreadsheet; open a new town only once a delivery route genuinely exists; set commission per category rather than one flat rate across electronics and apparel; run the storefront in the language the market actually reads.
Proposed configuration: cash on delivery with a server-verified code at handover, a cash-collected balance per agent settled to the operator wallet, an operator-set handling fee, zip allowlists deciding where checkout is offered, tax classes bound per zone, and commission per category.
Thailand - provincial and cross-border marketplace
What this market makes hard: delivery economics differing sharply between a dense capital route and a provincial one; buyers expecting to pay on arrival; neighbouring markets quoting in a different currency.
What an operator would want: serve provinces and capital from one catalog without one subsidising the other; quote every buyer in the currency they recognise; add a neighbouring market as configuration rather than a second platform; keep the last mile, and the cash, under the operator brand.
Proposed configuration: zones with per-category shipping cost overrides, a currency catalog whose default drives checkout display, cash on delivery with per-agent reconciliation, delivery checkpoints, a multi-language storefront across every surface, and nine order statuses with a full status history per order.
Why show models at all
We would rather show how the proposed platform configures for a market than borrow credibility from a deployment that has not happened. There is no review or rating markup anywhere in this bundle.
Frequently Asked Questions
Is the platform secure and is it certified?
The proposed controls are named: Argon2id hashing, CSRF protection, parameterized queries, HTML sanitization, tenant scoping, webhook signature verification and append-only audit ledgers. The design is intended to be ready for a penetration test and for ISO 27001 or SOC 2 evidence gathering. This preview supplies no ISO 27001 or SOC 2 evidence; no certification is claimed here.
Can you show me a client running this platform?
Not for this product, and we say so plainly. We have not yet deployed this multi-vendor commerce platform for a named client, so the reference configurations above are modelled, not engagements. We would rather walk you through the proposed scope in detail than borrow credibility from work that has not happened.
Do I get the source code?
Yes - that is the proposed handover. The full backend and all three Flutter projects transfer to your repository. Nothing is obfuscated, there is no licence callback, and you are not dependent on us to keep the platform running.
How is commission calculated?
Proposed: a percentage set globally, per category, per vendor or per product, deducted from the vendor wallet when the order completes and reversed automatically if the order is refunded. Vendor subscription tiers can carry reduced rates as a paid benefit.
Am I locked in to you after launch?
No, and the stack choice is deliberate. PHP, Laravel, MySQL, Redis and Flutter are conventional and hireable, so any competent team can take the platform over. The source transfers with reversible migrations, there is no licence callback, and the proposed commercial terms carry no per-order fee that would give us leverage over you later.
Does it support multiple currencies and languages?
Proposed: currencies carry operator-set exchange rates, a polymorphic translation store holds interface and content strings, and delivery country codes and zip allowlists control where you ship. Rates are operator-set rather than fetched automatically, which is deliberate - a marketplace that silently reprices surprises vendors and customers at the same time.
Independence and third-party names
Omnistu is an independent software development project. We are not affiliated with, connected to, sponsored by, or endorsed by Amazon. "Amazon clone" is used descriptively - it is how the software industry refers to building a platform with similar functionality, and how buyers search for it. Amazon and all other third-party names and marks are the property of their respective owners, referenced here solely to describe the category of software offered.
Discuss your scope by message
Tell us your commerce categories, your expected vendor count and your delivery model. We reply by email with a scoped proposal - no calls required, no phone numbers collected.
