What is this project? Kitrum re-architected and scaled Tawkify's matchmaking platform — a US-based personalized dating service operating since 2012 that connects clients through human matchmakers rather than algorithms. What problem did Kitrum solve? Tawkify's microservice platform across four applications was creating bottlenecks: matchmakers couldn't scale their caseload efficiently, and the time between a client registering and going on a first date was too long — directly hurting retention and revenue. How Kitrum built it Kitrum embedded as a long-term engineering partner, re-architecting the platform for scalability, redesigning the business logic for the matchmaking workflow, streamlining client onboarding, and building a QA department from scratch. Testing included nightly functional and integration tests, regression testing on every two-week release cycle, accessibility testing, and GraphQL security testing. The team operated on a classic Agile model with bi-weekly sprints aligned to US Pacific Time. Key technical capabilities delivered: - Full platform re-architecture across internal and customer-facing applications - End-to-end QA infrastructure built from zero: automated testing with Cypress (TypeScript), performance monitoring via TestRail, vulnerability assessments on GraphQL APIs - Streamlined onboarding flow reducing friction between registration and first match Results: - 20% reduction in time from client registration to first date - Unified performance monitoring process reusable across future teams - Ongoing engagement — Kitrum continues as active tech partner Tech stack: JavaScript, TypeScript, Node.js, PHP, Figma, TestRail, Cypress
KITRUM
RightFirms reviewed this company’s profile, work, reviews, clientele, and public trust signals.Engineering for live products where mistakes are expensive.
KITRUM Portfolios
What is this project? Kitrum built a two-stage subsea leak detection and predictive monitoring system for Franatech, a German manufacturer of underwater dissolved gas sensors used in offshore oil and gas, aquaculture, and deep-sea scientific research. What problem did Kitrum solve? Franatech needed software to visualize real-time sensor data and predict the location of underwater hydrocarbon leaks — deployable on a closed internal network with no internet access, stable enough to run for 5–10 years without upgrades or live support after deployment. How Kitrum built it Kitrum delivered in two stages. Stage one: an MVP for real-time sensor data visualization completed in 2 months, initially in Python, then with a React frontend. Stage two: a predictive leak monitoring system completed in 5 months, built in close collaboration with Franatech's partner research institute. The system was architected offline-first, containerized with Docker for portability, and stored using SQLite — chosen for its ability to handle hundreds of gigabytes of historical sensor data without network dependency. All libraries used carried free commercial licenses to support decade-long operation. Key technical capabilities delivered - Real-time visualization of sensor data across water temperature, pressure, and movement measurements from multiple source types - Predictive simulation engine allowing operators to model potential future leak locations - Offline-first architecture with local disk logging — no internet connection required post-deployment - Complex mathematical models for dissolved gas behavior integrated directly into the monitoring system Results - MVP delivered in 2 months - Full predictive monitoring system delivered in 5 months on schedule - Zero post-deployment connectivity required — fully self-contained on client's internal Windows Server network - Active detection and operator alerting for underwater oil and gas leak events Tech stack: Python, C#, React, TypeScript, SQLite, Docker
What is this project? Kitrum re-platformed a complex data science application for Quantium — a global AI and data science firm with 1,200+ specialists across 14 offices — migrating from a legacy technology stack to a modern, scalable architecture while maintaining full system continuity throughout. What problem did Kitrum solve? Quantium's existing platform carried accumulated technical debt: bugs affecting system stability, a direct database connection architecture limiting scalability, and a legacy stack requiring replacement without disrupting live operations. How Kitrum built it Kitrum operated via an outstaffing model embedded within Quantium's engineering team. The engagement ran in two phases. Phase one (~6 months): systematic issue identification using user feedback, support tickets, and internal testing; deep code review; bug reproduction and isolation; and solution implementation with full unit, integration, and end-to-end test coverage. Phase two (3.5 years): full re-platforming — refactoring the codebase, replacing direct database connections with a metrics library intermediary layer for improved error handling and scalability, integrating performance monitoring, and migrating to the new stack with comprehensive validation at every stage. Key technical capabilities delivered - Full codebase refactor and stack migration with zero loss of existing functionality - Metrics library integration replacing direct database architecture — enabling performance monitoring, faster debugging, and future-proof maintenance - Performance optimization based on collected metrics data: reduced database queries, faster response times, improved resource utilization - Comprehensive documentation maintained throughout for knowledge transfer Results - Seamless migration from legacy to modern stack across a 3.5-year engagement - Reduced operational costs through more efficient infrastructure and resource utilization - Improved system stability, maintainability, and scalability - Ongoing engagement — Kitrum remains active tech partner for Quantium Tech stack: .NET, Angular, AngularJS, RabbitMQ, Terraform
What is this project? Kitrum designed and built Lula's full online banking and funding platform — a South Africa-based fintech serving SMEs with bank accounts, prepaid cards, and three distinct funding products. After closing a $6.5M Series A, Lula partnered with Kitrum to scale the platform domestically and across Africa. What problem did Kitrum solve? Building a regulated fintech platform for Africa required solving infrastructure constraints that don't exist in Western markets: inconsistent internet connectivity requiring offline functionality, mobile-first device usage across multiple OS versions, compliance with local financial regulations, and integration of 7 third-party financial tools including Mastercard, identity verification, and transaction classification systems. How Kitrum built it Kitrum assembled a cross-functional team of ~50 specialists over the engagement, including backend engineers, frontend developers, QA, DevOps, solution architects, and PMs. The platform uses a layered architecture combining monolith, microservices, and serverless approaches — enabling faster feature development while maintaining enterprise-grade scalability. An MVP covering core banking, card management, and funding was delivered in 7 months. The platform was subsequently expanded with bulk payments, scheduled transfers, analytics via Strands integration, 2FA security, and 24/7 engineering support. Redis caching was implemented to improve speed and reliability under low-connectivity conditions. Kitrum is now building the companion mobile app in React Native. Key technical capabilities delivered - Full online banking system: account management, virtual and physical Mastercard issuance, inter-account transfers, bulk payments, scheduled/recurring payments - Three funding products: revolving capital facility, capital advance, and Lulapay - 7 third-party integrations: Mastercard PTS, Traderoot, Jumio (KYC/AML), Strands (analytics), Gorbank, AirShip, Stitch - Offline-capable lightweight architecture with local data center routing for low-connectivity environments - 24/7 engineering support post-launch Results - MVP delivered on time and within budget in 7 months - Up to $1,000/month in Azure cloud cost savings through infrastructure optimization - Platform supports three distinct SME funding products across South Africa - Ongoing engagement: web platform maintenance plus active React Native mobile app development Tech stack: React, React Native, JavaScript, C#, .NET, .NET Core, Microsoft Azure, AWS
What is GetMeo? GetMeo is a B2B AI knowledge agent built to solve corporate information overload. It creates a persistent memory layer for companies — connecting data from Slack, Notion, CRM systems, emails, documents, and web links into a single, searchable knowledge base that understands context, not just keywords. The problem it solves Teams lose hours weekly searching for information scattered across tools. Existing solutions like Glean target enterprise-scale deployments with no data control for smaller teams. GetMeo was built for project-level knowledge management with security and data ownership as a core requirement. How Kitrum built it Kitrum developed GetMeo using a proprietary HybridRAG architecture — combining VectorRAG for high-speed semantic search with GraphRAG for relationship mapping between data points. This resolves the most common AI knowledge management failure: losing context between queries. The system was built in 3 months by a lean team of four: product manager, full-stack developer, backend developer, and frontend developer. One month was spent in close collaboration with the core AI technology partner to stabilize and integrate the underlying model before two months of active product development. Key technical capability HybridRAG architecture achieves 70–80% better outcomes in real-world retrieval tests compared to single-method approaches. The system ingests data from emails, documents, YouTube videos, web links, and SaaS tools without requiring manual tagging or categorization. Results Fully functional B2B AI agent delivered in 3 months 70–80% improvement in knowledge retrieval accuracy vs. baseline Integrates with Slack, Notion, and CRM systems out of the box Active development continuing with expanded feature set Tech stack: AI/LLM integration, HybridRAG (VectorRAG + GraphRAG), Slack API, Notion API, CRM integrations
Turf Tank, an innovator in autonomous field painting, leveraged its expertise as KITRUM developed an Android app utilizing .NET, Flutter, and Azure.
Ledgerset, a pioneer in supply chain networks, leverages KITRUM's bespoke blockchain system, boosting efficiency for retailers and suppliers.
StackSource, a leader in real estate loans, along with KITRUM, migrated to React.js, improving UI and offer creation for optimized efficiency.