Product Description

Salik is a last-mile delivery marketplace. Food is the first vertical; the same rails are designed for any sender.

What Salik is — and isn't

Salik is not a food-ordering app for the public, and not a traditional courier employer. It is a technology platform that connects merchants who need delivery with independent couriers, and settles the money.

The four people in every delivery

Positioning

A general-purpose last-mile platform, food-first. The data model deliberately treats the “restaurant” as a generic merchant/sender, so the same product can later serve marketplace deliveries, business correspondence, and peer-to-peer parcels.

Brand & customer touchpoints

Sender brand is Salik, with the merchant's name in the message body. Salik runs a public brand/profile website. Customers aren't direct users today, but their feedback is captured (see Operations → reviews).

How It Works

Stage 0 — Onboarding 🟢 partial

Built: courier self-registration + the admin verification flow (approve / hold / reject / ban). Designed, not yet built: the eligibility gates and document capture.

• Couriers: basic registration + admin approval are built ✅. The 18+ gate, selfie-with-ID upload, driving-licence check (moped/car), thermobag flag, and optional operator WhatsApp video-verify marker are designed 📐. No exclusivity; courier provides their own thermal bag.
• Merchants: self-serve onboarding (legal name, address, contact, hours) + team verification is designed 📐. Merchants are currently provisioned by an admin.

Stage 1–2 — Order & auction 🟢 partial

• Customer orders from the merchant off-platform.
• Merchant posts the delivery — basic posting (address, payment method, notes, price) is built ✅. Structured address + Google Maps pin, size/vehicle, thermobag flag, and food-ready timer are designed 📐.
• Suggested starting price (distance × weight × urgency × weather; floor 3 AZN) — designed 📐 🔑 Maps + Weather.
• Auction (2–3 min sessions; rebid or cancel on no-accept) — the headline differentiator, designed but not yet built 📐. Today a published order is accepted first-come at the merchant's set price.

Stage 3–4 — Accept & pickup 🟢 partial

• Courier accepts a job and enters the 5-digit pickup code shown on the merchant screen — built ✅ (race-safe accept).
• Pre-accept distance/zone view and visibility filters (vehicle/size, thermobag, COD wallet balance) are designed 📐.
• Package credit consumed per delivery, with a hard block when a merchant is out of credits — built ✅.
• 15-min wait limit + waiting fee — designed 📐 (admin-configurable).

Stage 5 — Delivery 🟢 partial

• Courier enters the 5-digit delivery code to confirm hand-off — built ✅ (constant-time check + attempt lockout).
• Customer notification at pickup (ETA + courier phone + delivery code + rating link) over WhatsApp/SMS — designed 📐 🔑 WhatsApp; an SMS hook is stubbed.
• Leave-at-door, someone-else-receives, and mandatory drop-off photo (when no code) — designed 📐.
• Tips, 100% to courier — designed 📐.

Stage 6 — Settlement 🟢 partial 🔑 CIBPAY

• A double-entry ledger posts each delivery's earning (courier +, merchant −) atomically, with a daily zero-drift reconciliation check — built ✅.
• Wallet escrow / COD hold-and-release, courier float + security deposit, and gateway top-up/instant withdrawal — designed 📐 🔑 CIBPAY.
• Two review streams (in-house courier rating + optional Google Reviews deep-link) — designed 📐.

Technology Stack

A modular monolith designed so each module can later be extracted into a microservice — scaling is extraction, not a rewrite.

Stack

Modules

identity · orders · couriers · restaurants · wallet · notifications · partner-api · webhooks. Modules talk through public interfaces and an event outbox — never by reaching into each other's tables.

External integrations 🔑 provider decisions

Engineering conventions

• Conventional Commits; TypeScript strict (any is an error).
• Money is always Decimal — never float; renders as “X.XX AZN”.
• No defaulted secrets — services fail fast on missing env.
• Module boundaries enforced by dependency-cruiser in CI.

Build Status & Roadmap

Reflects the codebase + internal backlog as of 2026-06-15. “Built” items were verified running in the app; “designed” items are specced in the decision docs but not yet implemented.

✅ Built & running

• Modular-monolith backend + OpenAPI contracts + a web app for all three roles (admin, merchant, courier).
• Full order lifecycle — publish → accept → arrived → pickup → on-the-way → delivered, with cancel/dispute, race-safe accept, and 5-digit handoff codes (constant-time check + attempt lockout).
• Double-entry wallet ledger with per-delivery posting and a daily zero-drift reconciliation check (admin “Balanslar” view).
• Prepaid-package monetization — admin-editable tiers, credit grants, FIFO consumption, and a hard block when a merchant runs out.
• 8-page admin console — dashboard, couriers (with verification flow: approve / hold / reject / ban), merchants, orders + order detail (status timeline, assigned courier, linked dispute), disputes, ledger, settings.
• on_hold reversible courier status; courier & merchant detail pages; consistent loading/error states across the console.

📐 Designed & specced — next to build (pilot)

• Onboarding & identity gates (18+ , selfie-with-ID, licence, thermobag, video-verify marker; merchant self-serve onboarding).
• Order creation extras (structured address + Maps pin; size/vehicle, thermobag, food-ready timer tagging).
• Pricing engine (distance × weight × urgency × weather; floor 3 AZN).
• Auction (2–3 min sessions, rebid/cancel, distance-before-accept).
• Batching capability — build, then feature-flag off at launch (no later migration).
• Delivery flow (notifications, leave-at-door, drop-off photo, wait + waiting fee).
• Two review streams (in-house courier rating + optional Google Reviews).
• Wallet & settlement (COD escrow, 10 AZN deposit, tips, gateway top-up/withdrawal, admin credit line).
• Dispute-rules engine, role-based access (RBAC), audit log, exportable delivery log.
• Ops (hours window, zones, image retention 3 days → archive).

🔑 Blocked on external providers

• CIBPAY, WhatsApp BSP, Google Maps API, Weather API.

⏳ Awaiting lawyer (pre-launch blockers)

• Insurance & accident liability; courier + merchant agreements; ToS; privacy policy; customer notice wording; payment/e-money & data-retention review.

🔮 Future phases

• P2P delivery (customer as sender) · live order-tracking map · SIMA / SIMA İmza · driver fine-history check · public brand site · online split-payment link.

Product in Action

Real screenshots from the running app — not mockups. Captured 15 Jun 2026.

Admin dashboard

Overview charts, order volume, and quick stats visible to the operator.

Double-entry ledger (Balanslar)

Per-delivery postings with courier names, AZN amounts, and live reconciliation. Drift shown as 0.00 AZN.

Order detail — status timeline + dispute panel

Admin order detail page: assigned courier cross-link, full status timeline, and inline dispute resolution.

Courier board (mobile)

Live job board on the courier's phone — 3 open deliveries with price, district, and accept button.

Merchant credits view

Merchant sees their active package, remaining credits, and usage history.

Intellectual Property

Items to formalize:

• Brand: “Salik” name & logo — trademark registration in Azerbaijan (to do).
• Domain & social handles — secure and document.
• Codebase ownership — greenfield build; confirm assignment/ownership in founder & contractor agreements.
• Proprietary logic — the auction + pricing-engine design as a know-how asset.

Market Research

Market context

Azerbaijan’s food delivery market is dominated by Wolt, Bolt Food, and Yandex Food — all aggregators that own the customer relationship and charge 25–35% commission on food revenue. Thousands of restaurants in Baku operate outside these platforms entirely, taking orders via phone and Instagram and running informal self-delivery.

Who Salik targets

Restaurants and cafés that already have order demand and need affordable delivery — not discoverability. Specifically:

• Establishments currently self-delivering (managing drivers ad-hoc, high ops burden).
• Aggregator-listed merchants paying 25–35% commission and wanting a cheaper overflow channel.
• Instagram-only businesses taking orders via DM with no delivery infrastructure.

Why the timing is right

• Wolt/Bolt expanded to Azerbaijan but focus on high-volume anchor restaurants — mid-tier and neighbourhood venues are underserved.
• WhatsApp Business adoption is high among restaurant owners — the referral channel is essentially free.
• No neutral last-mile-only marketplace exists in the market today.

Still to research (formally)

• Exact restaurant/café count in Baku, Absheron, Sumgayit pilot districts.
• Courier supply pool size and prevailing daily earnings.
• Average basket size, order frequency, seasonality.
• Regulatory environment for gig/delivery work in Azerbaijan.

Competitive Analysis

Landscape

Our edge

• Lower cost for merchants who already have demand and just need delivery.
• Merchants keep their own customers and brand.
• Auction balances supply/price dynamically.

To research

• Exact competitor pricing, coverage, courier counts, and merchant sentiment in the pilot districts.

Revenue Model

Revenue comes from prepaid delivery packages sold to merchants — not a commission on their food sales.

Packages ✅

• A credit is consumed on pickup. Blended revenue ≈ ~0.90–1.00 AZN/delivery depending on package mix.
• Credits run out → hard block; super-admin may grant an admin-approved credit line (limit, visible “owed”, netted on next purchase).

Wallet & COD 🟢 partial

The double-entry ledger is built ✅. COD escrow (block on the courier wallet at pickup, release to the merchant on delivery) is designed 📐 🔑 CIBPAY. Future revenue lines: withdrawal fees, optional per-COD platform fee, online split-payment.

Charge from launch

Salik charges from day one; acquisition cost is offset with referral credits and coupons rather than running free.

Pricing & Auction

The auction 📐 designed

The merchant sets/bids a delivery price; couriers compete to accept within a 2–3 minute session. On no-accept, the merchant rebids (new session) or cancels. The platform never auto-raises the price — the market does.

Suggested price 📐 designed 🔑 Maps + Weather

A pricing engine suggests a starting price from:

• Distance (Google Maps Distance Matrix)
• Weight / size
• Urgency
• Weather (surge on rain/snow/extreme conditions)

Floor price 3 AZN. The merchant can always bid above the suggestion.

Other fees

• Waiting fee: merchant can pay ~1 AZN per 10 min when a courier waits (admin-configurable).
• Tips: 100% to the courier.
• Security deposit: 10 AZN default (admin-configurable) to access COD orders.

Customer Segments

Merchants (paying customers)

• Restaurants, cafes, and shops that already have order demand and need delivery.
• Especially those avoiding high aggregator commissions or running ad-hoc self-delivery.

Couriers (supply)

• Independent couriers (bicycle / moped / car), non-exclusive.
• Motivated by fair auction pay, instant withdrawal, and gamified incentives.

End customers (recipients)

• Not direct users today; reached via WhatsApp/SMS.
• 🔮 Future P2P phase turns customers into senders.

Go-To-Market

District-by-district launch, seeded by anchor merchants and grown through referrals.

Approach

1. Anchor merchants first: personally onboard ~5 high-volume cafes per district as VIPs — they become the first referrers and create courier demand.
2. Seed courier supply: recruit couriers ahead of demand with sign-up bonuses + gamification so the auction always has takers.
3. Local launch moment: a district-specific coupon distributed via supplier WhatsApp groups makes each launch feel like an event.
4. Referral flywheel: merchants refer merchants for free deliveries (see Referral & Coupons).
5. Expand: Baku → Absheron → Sumgayit as supply/demand stabilize.

Channels

• WhatsApp forwards between restaurant owners (highest-leverage in Azerbaijan).
• Food-supplier network partnerships (co-branded coupons, zero cash cost).
• Instagram/social seeding via top merchants.

Referral & Coupons

Core mechanic

Every active merchant gets a unique referral code. New merchant signs up with it → referrer earns 20 free delivery credits, new merchant gets 10 (credited on first paid delivery to prevent fakes).

Tiers

Launch coupons

• Welcome codes at onboarding (e.g. 10 free deliveries; VIP 25).
• District-launch public code for the first merchant from a new area.
• Supplier-network co-branded codes.

Full detail: Referral & Marketing Strategy document.

Courier Gamification

Supply (couriers online) is the scarce input in an auction marketplace. Gamification keeps couriers active and loyal — mostly at zero cash cost.

Mechanics

• XP & levels: earn XP per delivery, 5-star ratings, peak-hour presence, on-time pickups. Levels unlock perks — the strongest being a few-second head-start on high-value orders for top couriers (zero platform cost).
• Streaks: daily-delivery streaks with bonuses; loyalty “freeze” for higher levels.
• Missions: admin-set, time-boxed challenges that pull supply to gaps (peak hours, late nights, bad weather, new districts).
• Leaderboards: weekly, district-scoped so newcomers can compete.
• Badges: recognition, zero cost.
• Cold-start cash: sign-up bonus, refer-a-courier, optional launch-week earnings guarantee — time-boxed and budget-capped.

Post-MVP. Full detail: Gamification Strategy document.

Onboarding Pipeline

Merchant pipeline

Courier pipeline

Unit Economics & Costs

Per-delivery cost drivers

• SMS OTP ≈ 0.03 AZN (3 qapik) per confirmation (WhatsApp messaging priced separately as primary channel).
• Payment gateway (CIBPAY) fees on top-ups & withdrawals.
• Google Maps API call per pricing/address lookup; Weather API (free tier).
• Courier incentives/gamification allocation (budget-capped).

Acquisition & fixed costs

• Referral cost ≈ 3 AZN per acquired merchant.
• Support/CRM subscription (respond.io / Wati); ops staffing.
• Hosting (Hostinger VPS).

Float (not revenue)

• Courier security deposits (10 AZN each) and wallet balances.

Projections

Revenue mechanics

Salik earns on prepaid package credits consumed at pickup — not on the delivery fee, which is a merchant→courier pass-through. Blended revenue ≈ ~1 AZN per credit (package mix-weighted). Secondary income: breakage on expired Trial/Main credits (Premium rolls over).

Unit economics (Base)

Break-even

Break-even assumes ~70 deliveries/merchant/month (Base). The Conservative scenario loses 24.7k AZN/yr; Aggressive earns 209k AZN/yr.

12-month ramp

Month-by-month P&L is in the model workbook (work/Salik_Financial_Model.xlsx), with a pilot ramp starting from 0 merchants, reaching break-even by month 4 in the Base scenario.

What will move the model most

• Active merchant count — the primary lever; 94 is achievable with one district launch.
• Deliveries per merchant — assumed 70/mo; restaurants with self-delivery often do 100–200/mo.
• Package mix — higher Premium uptake improves blended revenue/credit (Premium: 200 credits, rollover).
• Gateway costs — CIBPAY rate is the largest unconfirmed variable.

Operations & Support

Hours ✅

Orders accepted 08:00–02:00; deliveries complete until 03:00.

Geography ✅

Baku, Absheron, Sumgayit. Distance-based pricing.

Support 🔑 CRM

Business WhatsApp integrated into a CRM (respond.io / Wati suggested), Azerbaijani only, staffed by platform support.

Reviews

• Courier/delivery rating — in-house, Salik-owned, feeds gamification.
• Restaurant review — optional Google Reviews deep-link (posts to the merchant's real Google profile; goodwill only).

Data

Salik retains operational data; delivery images kept 3 days then archived; courier can see the delivery address.

Dispute Resolution

Principles

• Courier not at fault is always paid by the merchant.
• Merchant may gift the food to the courier instead of a return trip; return-to-sender is a paid trip.
• No delivery-fee refund; food quality is a merchant↔customer matter.
• The pickup + delivery codes and wallet escrow are the evidence backbone.

Flow

Rule sets cover cancellations, failed deliveries, food quality, COD, and a small platform-compensation budget. Resolution target: ASAP.

Admin & Access (RBAC)

Roles

Least-privilege, separation of duties, full audit log on every privileged action.

Courier conduct

Tiered enforcement: warning → temporary hold (on_hold) → permanent ban. Bans reserved for theft/fraud, violence, deliberate code bypass, or repeat serious violations.

Full detail: Admin Access Framework document.

Agreements & Contracts

Deliverables requested from counsel

• Courier Agreement (independent contractor; liability; deposit/wallet; conduct).
• Merchant Agreement (marketplace terms; no delivery SLA; package terms; data consent).
• Platform Terms of Service & Privacy Policy.
• End-customer notice wording for the WhatsApp/SMS message.

Legal opinions requested

• Insurance & accident liability structure.
• Customer data & privacy (lawful basis, retention, courier access).
• Payments / e-money licensing (is CIBPAY sufficient?).
• COD & tax/fiscal-receipt obligations.
• Enforceability of dispute rules & no-refund terms.

Full detail: Lawyer Brief document.

Regulatory Compliance

• Payments / e-money: assume the CIBPAY partnership avoids a separate licence — to confirm.
• Data protection: retention policy and lawful basis for messaging customers (merchant-provided numbers) — to confirm.
• Gig / labour: courier classified as independent contractor — to confirm.
• Consumer protection: extent of obligations given the customer is not our direct contracting party — to confirm.
• Tax: VAT / fiscal receipts on packages and delivery fees — to confirm.

Risks & Open Items

Pre-launch blockers

• 🔑 Provider integrations: CIBPAY, WhatsApp BSP, Google Maps, Weather.
• ⏳ Legal: agreements, insurance/liability, ToS/privacy.

Market / execution risks

• Cold-start: no couriers online → auction fails. Mitigation: seed supply + gamification + sign-up bonuses.
• Auction no-accepts: mitigated by rebid flow + price guidance.
• Trust on first deliveries: code system, wallet escrow, drop-off photos.
• Aggregator competition: differentiate on cost + merchant ownership of customers.

Research still to do

• 📋 Market sizing, competitor data, regulatory confirmation, financial projections.

Source Documents

Internal working documents this room is built from (in ../work/):

• Salik_PRD.md — product requirements (authoritative).
• Open_questions.md — all 60 business decisions.
• Salik_Implementation_Plan.md — build workstreams & sequencing.
• Salik_Backlog.md — status & future-phase ideas.
• Dispute_Rules_Framework.md — dispute logic (lawyer review pending).
• Admin_Access_Framework.md — RBAC + conduct policy.
• Referral_Marketing_Strategy.md — merchant acquisition.
• Salik_Gamification_Strategy.md — courier engagement.
• Lawyer_Brief.md — legal engagement brief.

Glossary