Evidence before advice
Recorded cash differences, stock pressure and cost changes support the report. A cash difference is not automatically theft; delayed orders are not automatically lost sales.
Product & founder portfolioBatam, Indonesia · 2025–present
INDOWYSS
One operating ecosystem.
Four connected advantages.
A business should get more from its daily records than another dashboard. INDOWYSS connects an automated Manager, shared promotion, collective purchasing and modern operations so owners can understand what is happening, decide what to do and work with other businesses.
Stage Pre-revenueBuilt by Independent founderEvidence Working software & demo records
01 / The ecosystem
Not four separate products.
Transactions, costs and purchases create the operating record. The Manager uses that record; Cross Promo creates a route to other customers; purchasing connects compatible needs with supplier offers. The value is in the connection.
01 / Automated Manager
Know what needs attention.
See the reason behind the recommendation.
The Manager is an automated system, not a person replacing the owner. Hari Ini separates immediate work from Rencana Usaha, where the owner can review changes to prices, staffing or equipment.
Owner control
A priority should lead to an action.
Daily tasks and business plans have separate places. Evidence stays attached to the decision, with a route back to the operational record.
01 / 04
Simulate before spending
Use recorded menu costs and your own estimates to compare extra sales contribution with staff or equipment costs. The result is conditional, not a promise of profit.
Follow up after acting
Record a trial and compare later operating records. A better period is evidence to review, not proof that a single action caused the improvement.
How a report is presented
Web: evidence → calculation or scenario when needed → decision. Technical detail and assumptions stay available without overwhelming the summary.
WhatsApp: a short, actionable summary only when attention is needed, subject to consent, channel availability and repeat-message controls.
Real WhatsApp delivery and paid external AI are not active in the demonstrated environment. The current Manager already uses backend calculations and coded rules.
02 / Cross Promo
A customer at another outlet
can discover your business.
Owners choose the promotion and its conditions. The offer appears on receipts from participating outlets, with the destination and a claim path. It is shared customer reach, not a promise that every recipient will visit.
From one purchase to a new introduction
- A customer shops at outlet A.
An ordinary sale creates a receipt.
- The receipt carries outlet B's promotion.
For example, a free fruit soup under B's chosen terms.
- The customer follows the claim path.
The destination, conditions and voucher/QR give the offer a concrete next step.
The network is the product value. An outlet can introduce itself beyond its own existing customers. Conversion and repeat visits still need to be measured.
03 / Collective purchasing & suppliers
Small needs can add up
to a supplier's minimum order.
An outlet chooses its own quantity. Compatible needs are combined by item, specification, area and fulfilment terms. Suppliers can contribute offers; buyers decide only after seeing the final terms.
Outlet needs
Interest is the beginning, not an order.
Each need shows an item, area, specification and quantity. Posting interest does not create a payment obligation.
01 / 03
- 01Post a needYour item and quantity
- 02Review the offerPrice, minimum, quality and location
- 03Commit and payOnly after terms are accepted
- 04Receive the goodsAgreed pickup or delivery
- 05Keep the resultOrder history and recorded price
A workspace for both sides
Buyers track interest, orders, progress and history. Suppliers review stock confirmation, offers and preparation work. Demand and a final order remain different records.
Terms must actually match
A buyer requesting delivery is not silently matched to pickup-only fulfilment. Price, location, schedule, costs and fulfilment need explicit agreement.
Payment stages, fees and current limits
The implemented model separates goods value, buyer fee, supplier deduction and partner charges. New agreements use a 0.5% buyer fee and 0.5% supplier fee; previous accepted terms are preserved.
Under this model, the platform fee is earned on received goods. Failed payment-MOQ rounds refund the paid amount, including the buyer fee. These are tested sandbox rules, not a claim of active money movement.
Production payment, refund and payout require partner activation and real end-to-end verification. Bank/payment-partner charges are separate.
04 / Modern operations
Daily work feeds
the same operating record.
Browser-based transactions, receipts, shifts, expenses, closing, stock and menu costs form the foundation. The goal is practical daily work on a phone, tablet or computer, with no app download.
Restaurant Beta
Extend the record, not a disconnected system.
Tables, orders and kitchen/billing states build on the existing POS record in an isolated beta lane.
01 / 02
Role-aware workflows
Owner decisions and staff operations have different permissions. Sensitive price, stock, refund and financial changes require the authorised role.
Costs that can be inspected
Menu costs can separate ingredients, direct costs and allocated overhead. Missing cost coverage limits the conclusion; revenue is not labelled net profit.
Failure paths matter
Local tests cover repeated refunds, concurrent closing, sync retries and queue failure. Those tests support delivery discipline, not a blanket production guarantee.
02 / INDOWYSS OS & AI
From a record to a useful decision.
The OS direction is broader than a cash register: a common discipline for data quality, calculations, decisions, permissions and follow-up across sources and business types. The current Manager is grounded in POS records; the universal multi-source path is still being developed.
- 01Collect
Transactions, closing, purchases, costs and authorised operational inputs.
- 02Validate & scope
Check inputs, ownership, units, time and duplicate records.
- 03Check evidence
Use available history and expose missing or weak data.
- 04Calculate
Backend calculations produce figures and explicit assumptions.
- 05Explain & prioritise
Rules select attention and actions; AI can assist with explanation.
- 06Act & evaluate
The owner approves important actions, then reviews later records.
Backend calculates. AI helps explain. The owner decides. An AI model must not invent costs, savings or investment returns. It must not change prices, issue refunds or move money on its own.
What can improve with more data?
Longer outlet history, more complete costs and recorded outcomes can give future analysis stronger context. Anonymised cross-outlet benchmarks are a direction subject to consent and privacy thresholds, not permission to mix customer records.
What does not happen automatically?
Feedback does not automatically retrain an external model or prove a recommendation worked. External model integration is currently evaluated internally, not used to write owner reports automatically. Paid model execution is not active in the demonstrated environment.
03 / Development & evidence
A working product.
A deliberate commercial rollout.
Screenshots show application or receipt-format demonstrations, not customer financial outcomes. Core workflows have local test evidence; deployment and commercial validation are separate milestones.
- Built & locally validated
- POS records, calculation/rule-based Manager, costs, decision scenarios, action evaluations, Cross Promo and buyer/supplier workflows.
- Restaurant Beta
- Tables, bills, split workflows and kitchen/waiter coordination extend existing records. Restaurant-specific completion and production validation remain ongoing.
- Production activation ahead
- Paid external AI, real WhatsApp delivery and production partner payment flows are not active in the demonstrated environment.
- Commercial validation ahead
- No paying outlets as confirmed by the founder on 13 October 2026. Retention, measured savings and conversion are not claimed.
Financial integrity
Local regression testing has covered concurrent closing, repeated refunds, historical cost snapshots and server-side totals.
Operational resilience
Queue-outage behaviour, retry/deduplication and access boundaries are tested alongside responsive browser workflows.
Where the product began
In 2025, the founder worked with approximately ten small F&B operators in Batam for about one month. Paper records were parsed into condensed data, analysed with an AI agent and returned to the users.
This is founder-reported exploration. Original records are no longer available; it is not audited traction or proof of financial improvement.
The next stages
- Prove the first operating loop
Activate production services and validate daily owner/staff use.
- Measure the customer result
Track activation, retention, report usefulness and network conversion.
- Strengthen the network
Validate supplier terms, fulfilment and repeat purchasing.
- Expand only with evidence
Complete Restaurant Beta; later consider retail and other business types.
04 / Investor & partner context
The opportunity is an operating relationship,
not another isolated tool.
The starting customer is a small F&B owner who needs practical daily operations and clearer decisions. The long-term thesis is that useful records, owner follow-up and business-to-business coordination can reinforce each other.
Subscription model
Rp199,000 / outlet / month
One proposed subscription for the Manager ecosystem, including daily POS operations. This is the commercial offer, not realised recurring revenue.
Successful purchasing
1% combined platform fee
For new agreements: 0.5% from the buyer and 0.5% from the supplier, earned on received goods. Bank/partner fees are separate; real payment activation is still ahead.
The defensibility hypothesis
Evidence-linked decisions, recorded outcomes and repeated network participation may make the ecosystem more useful over time. Network effects and switching costs must be demonstrated; they are not established moats yet.
A controlled first cohort
The planned onboarding cap is 15 new outlets per week, with issues reviewed before expansion. It is a support and validation plan, not a current subscriber count.
What support would help next?
Distribution and marketing come first, alongside production service activation, incorporation and a focused team. The founder is open to aligned investors, mentors and commercial partners; no valuation, return guarantee or fixed fundraising target is asserted here.
What should be measured before scaling?
- Outlets that complete onboarding and use the records consistently.
- Paid conversion, retention and actual service cost per active outlet.
- Whether owners understand and use a recommendation, with later evidence.
- Promo claims and repeat purchasing, without assuming every impression converts.
- Supplier fulfilment quality, disputes and refund handling.
05 / Founder & product craft
Wisnu Falah
Independent founder, product and systems builder.
Based in Batam, Indonesia. INDOWYSS grew from field work and manual records into an AI-assisted software build. I define the product direction, workflows, evidence requirements and acceptance criteria, then work through implementation, testing and iteration.
- Product discovery
- Turn everyday owner and staff problems into usable requirements and a focused build sequence.
- Workflow & system design
- Connect roles, data, approvals, financial states and failure paths across the ecosystem.
- AI-assisted delivery
- Build frontend, backend and database workflows; inspect implementation and validate important behaviour.
- Communication & experience
- Refine plain-language reports, progressive detail and owner control so the system is understandable.
- Quality & iteration
- Use unit, database/API and browser checks, with scoped changes and documented limits.
Stack usedTypeScript · Next.js · NestJS · PostgreSQL · Prisma · Git
Practical backgroundFamily retail operations · F&B service · Sales · English teaching
Other workCommitTree is a separate side project in development, not a shipped or revenue-generating product.
Long-term vision: a trusted operating-intelligence foundation across business types, with specialised workflows and human control. Retail, workshops, showrooms, laundry and enterprise use remain future directions, not currently available product claims.
06 / Continue the conversation
For investors, partners
and people building useful businesses.
General product briefProduct, evidence, business model and next steps · PDFRecruitment portfolioPreserved application version, CV and project document
Wisnu Falahfalahwisnu711@gmail.com+62 819 1024 9645
Public portfolio only. Internal source code, credentials and customer records are excluded. Detailed due diligence can be shared separately with an agreed scope.