Developed an AI-native Learning Management System (LMS) leveraging LLMs and LangGraph to deliver engaging, interactive courses incorporating games, assessments, and multimedia content. Took an agentic approach to course creation using stateful graph-based agents for dynamic content generation and user interaction flows. Built scalable backend services using Go (Gin) and managed frontend interfaces using Remix and TypeScript. Provisioned and maintained cloud infrastructure with Pulumi, Docker, and Kubernetes, enabling reproducible environments and streamlined deployments.

Led the transformation of electronic invoicing into a SaaS business model, implementing self-serving client registration, tiered payment plans, and automated infrastructure provisioning. Engineered a data migration CLI tool in Rust, facilitating seamless transfer of legacy electronic invoice data with zero downtime. Developed services in Node/Go to generate shadow traffic, reducing system migration risks and allowing for gradual user adaptation.

Spearheaded a complete rewrite of the company's largest product, resulting in: 75% reduction in unplanned work, turnaround time improvement from days to hours, up to 30% reduction in infrastructure-related costs. Implemented high-performance Go/Rust services, gRPC for microservice communication, and event-driven architecture using Kafka. Created an efficient intranet back office using Next.js and tRPC, automating time-consuming processes and improving operational efficiency.

Designed and led development for a nation-wide project streamlining tax-related invoices and town hall administration processes. Increased 'Branchless Banking Agent' solution performance by 70% in latency reduction and 50% in resource usage by rewriting the core in Rust. Improved team turnaround time by 40% through introduction of DevOps CI/CD and trunk-based development practices.

Led development of a mobile app for a major local bank, streamlining invoice financing for corporate clients. Designed and developed a 'Branchless Banking Agent' solution, enabling over 500 local businesses to offer basic banking services. Increased utility margin on key hardware sales by 50% through in-house SDK development.

Developed a CRM system leveraging ML models to prioritize company project efforts, utilizing Python (pandas, scikit-learn, NLP, TensorFlow) and Angular.