Our Portfolio
Real Products.
Real Results. In Production.
Every project below is live — serving real users, generating real revenue, and engineered to scale without breaking.
Oh! Ride — Sri Lanka’s First SaaS Ride-Hailing Platform
OXHorn came to us with an idea that challenged the entire ride-hailing industry: what if drivers kept 100% of every fare, and the platform made money through a flat monthly subscription instead of commissions? We turned that idea into a production-grade, multi-city SaaS platform — live for 2+ years with zero major outages.
Build a real-time dispatch system that rivals Uber’s latency (<200ms GPS updates) using an 11-person team — while simultaneously developing a KYC pipeline that auto-approves drivers in 90 seconds using computer vision and OCR, across 8 independently deployable microservices.
Mave — B2B Chat with Automatic Cashflow Intelligence
Mave started with a simple observation: accountants already describe financial events in natural language when they chat with their team. “Invoice #204 paid — £1,200.” “Refund sent.” “Deposit received.” Mave listens to these messages, detects the financial intent using NLP, and logs it as a cashflow entry — automatically, without a second form or duplicate effort. The result: real-time communication and cashflow tracking unified into a single platform, replacing three separate tools. Delivered from design to production in just 8 weeks.
NLP-powered financial intent detection embedded directly in a real-time chat interface — so accountants never have to switch context between communication and bookkeeping.
Servicemate — B2B Service Management Platform
Field service businesses — plumbers, electricians, HVAC companies, cleaning services — were managing their entire operation across five completely disconnected tools: a calendar for jobs, a WhatsApp group for dispatch, a spreadsheet for tracking, a separate invoicing tool, and a reporting dashboard. Servicemate replaced all five with a single, unified SaaS platform. When a job is completed on the mobile app, an invoice is generated and sent automatically. Real-time job status updates sync across web and mobile in under one second. Business owners report saving 2–3 hours per day on administrative overhead alone.
Event-driven job lifecycle management: job completion on mobile triggers automatic invoice generation, payment tracking, and reporting updates — all in real time, across web and mobile simultaneously.
T&K — Custom E-Commerce & Inventory Management Platform
T&K is a retail business that outgrew the capabilities of off-the-shelf e-commerce solutions. Generic platforms like Shopify and WooCommerce couldn’t accommodate their specific inventory management requirements, multi-channel selling rules, and custom fulfilment workflows. We built a fully custom e-commerce platform — purpose-built for their exact operational model — with a real-time inventory system that syncs across physical stores and online channels simultaneously, preventing overselling and stockout scenarios that had been costing the business significant revenue.
Real-time multi-channel inventory synchronisation with conflict-resolution logic — ensuring stock levels remain accurate across physical and digital channels even during simultaneous high-demand events.
Why Our Code Holds Up
at Scale
The architectural decisions that make the difference between software that cracks under load and software that scales.
Every project we ship is built with an explicit assumption: this product will need to handle 10× the initial load within 18 months. That assumption changes how we design systems, choose databases, structure services, and write code from day one.
We don’t build monoliths that need to be rewritten when you grow. We don’t cut corners on database indexing because “it’s just an MVP.” We don’t skip error handling because “we’ll add it later.” Every layer of every product we deliver is engineered for the long term — because our reputation depends on it staying up.
Oh! Ride has run for 2+ years with 99.9%+ uptime. Servicemate handles real-time job dispatch across multiple concurrent sessions with sub-second latency. Mave processes financial events with NLP accuracy that doesn’t degrade under load. That’s the standard we hold everything to.
⚡ Microservices by Default
Independent services mean independent deployments, independent scaling, and isolated failures. One service going down doesn’t take the whole product with it.
🔄 Event-Driven Architecture
Apache Kafka or WebSocket-based event streaming ensures real-time data propagation with resilience — even under high-throughput conditions with temporary service interruptions.
🗄️ Database-per-Service Pattern
Each microservice owns its own data store — the right database for each job. PostgreSQL for relational data, MongoDB for documents, Redis for caching and geospatial queries.
☁️ Cloud-Native from Day One
AWS EKS with Kubernetes orchestration, auto-scaling groups, infrastructure-as-code, and zero-downtime rolling deployments — built in from the start, not retrofitted later.
Ready to Add Your
Product to This List?
Book a free 30-minute call and let’s scope out your project. No obligation — just a clear plan for what we’d build together and how long it would take.