Horeka.
The full platform architecture, order allocation algorithm, and WooCommerce-based delivery and discovery platform. Horeka Quick handles delivery via a weighted multi-factor algorithm. Horeka Discover handles in-store reservation. Store managers operate entirely via WhatsApp.
We needed a platform that connects customers to their neighbourhood retailers — not a dark store model, a retailer-first one — with intelligent order routing and zero learning curve for store managers.
— Client briefing · Horekaa Technocon Private Limited
The story
The context.
Horeka is not a dark-store model. It does not operate warehouses. It is a retailer-centric quick commerce platform — built on a simple but defensible insight: every neighbourhood already has 2–3 trusted home goods retailers. The logistics problem isn't storage, it's connection and coordination.
The business behind Horeka carries 50+ years of home goods retail industry experience. The technology layer — designed, architected, and built by Kyrex — takes that institutional knowledge and makes it operable at digital scale.
Kyrex's role here goes beyond build. Kyrex designed the order allocation algorithm, architected the full system, specified the delivery and discovery models, and continues to manage the server and platform infrastructure. This is a principal architecture and engineering engagement — not a web build.
What we built
02 workstreamsDeliverables.
Horeka Quick — delivery platform
Customers browse a unified product catalogue across all onboarded retailers in their pin code. They build a cart. Kyrex's order allocation algorithm determines which retailer(s) fulfil the order — and handles delivery via Porter.
01 — Platform architecture
| Component | Detail |
|---|---|
| Frontend | WordPress + WooCommerce — custom-themed, pin code based catalogue filtering |
| Backend | Custom PHP plugins — order logic, retailer management, allocation engine |
| CDN & Security | Cloudflare — performance, DDoS protection, edge caching |
| Infrastructure | Kyrex-managed VPS — server, security, backups, maintenance |
| Payments | UPI, Visa, Mastercard, Rupay, COD, Amex |
| Free shipping | Automated threshold — triggered above ₹1,499 |
02 — The order allocation algorithm
The most technically significant piece of engineering in the Horeka platform. Designed personally by Kyrex's founder — a weighted, multi-factor algorithm that determines which retailer fulfils a given order, handling complexity that no off-the-shelf plugin addresses.
| Factor | How It Works |
|---|---|
| Factor 1 — Distance | Real-time distance from each retailer to the delivery address, sourced via Google Maps API. Weighted as a primary factor. |
| Factor 2 — Delivery time | Estimated delivery duration from Porter, based on distance and current conditions. |
| Factor 3 — Delivery cost | Porter quotation for each potential assignment. Factored into the allocation decision to minimise cost. |
| Factor 4 — Product availability | Whether the retailer stocks all items in the cart. Availability is a primary weight — an unavailable retailer cannot be assigned. |
| Order splitting | When no single retailer can fulfil the complete cart, the algorithm automatically splits the order across multiple retailers — then merges delivery where possible. |
03 — Retailer acceptance via WhatsApp
Store managers do not use a portal or app. As soon as an order is allocated, the assigned retailer receives an instant WhatsApp message with accept/deny options and the ability to customise product quantities available. Zero learning curve for store managers.
04 — Porter delivery integration
Once an order is accepted by a retailer, it is passed to Porter for last-mile delivery. Order status updates flow back to the platform automatically via Porter webhooks — no manual tracking required.
Horeka Discover — in-store reservation platform
Home appliances — mixers, air purifiers, kitchen equipment — are high-consideration purchases. Many buyers want to verify quality, check the physical product, and compare options in person before committing. Discover solves the friction of not knowing what's available at which nearby store before making the trip.
| Step | What Happens |
|---|---|
| Discovery | Customer enters pin code — sees all nearby onboarded retailers and their available catalogue |
| Discover cart | Customer adds products to a Discover cart — not a purchase, a reservation intent |
| Reserve | Customer clicks Reserve — store manager receives WhatsApp notification instantly |
| Manager response | Accept, deny, or customise — same WhatsApp-based flow as Horeka Quick |
| Confirmation | Customer receives confirmation — their selections are reserved and held at the store |
| In-store visit | Customer visits knowing exactly what is available and reserved for them |
Discover does not force online purchasing behaviour where it does not naturally exist. It removes the single biggest barrier to in-store visits for considered purchases: uncertainty about availability.
Like what you saw?
Build the next one.
Two ways to begin. Both end the same way — a system in production, operated by Kyrex. Choose the entry that fits where you are.
Start a project.
Bring us the scope. We sharpen it, stand up the team, and you see the first working build inside three weeks.
- Response
- < 1 business day
- First build
- Week 3 · staging live
- Engagement
- 4 – 12 weeks
- Starts at
- Scoped to deliverables
Get the Blueprint.
A fixed-fee paid engagement. We map your operations, evaluate every realistic solution, and hand you a document you can act on — with us, or without us.
- Format
- Written report — yours to keep
- Duration
- 3 business days
- Outcome
- 3 paths · honest trade-offs
- Fee
- ₹15,000 · credited if you proceed