- Built the operator console and live 3D GIS common operating picture with React and Cesium (Blender-modelled 3D assets) on Go (Echo) services, smoothly rendering 1,000+ live aerial tracks.

- Designed the software architecture connecting sensors, embedded devices, backend services and effectors into one unified C2 interface for real-time monitoring, coordination and system control.

- Developed real-time data pipelines over REST, WebSocket, gRPC and MQTT, ingesting 10,000+ messages/sec across the C2 system.

- Contributed to the AI Fusion pipeline and Drone Detection module, correlating AI-generated targets with multi-sensor data for identification, threat assessment and tracking.

- Implemented C/C++ and Python hardware interfaces for radars, RF sensors, SDRs, EO/IR trackers and jammers, processing high-frequency telemetry with sub-10 ms latency.

- Integrated Echodyne EchoShield and EchoGuard radars and the Marduk Piraya passive EO/IR tracker, alongside ST Engineering's own S1000 AI/ML RF drone decoder, into the C2 track picture.

- Built interoperability services around the SAPIENT open standard (BSI Flex 335), so heterogeneous third-party sensors can plug into the C2 system.

- Developed SDR pipelines for I/Q signal acquisition, digital signal processing, demodulation and protocol analysis.

- Worked with RF link characteristics, including packet structures, synchronisation and frequency-hopping behaviour, to enable targeted, protocol-aware interference.

- Tackled system-level challenges in full-duplex RF operation: TX/RX isolation, filtering and self-interference management.

- Built the software integration between RF hardware, embedded systems and higher-level C2 control applications.

- Built the analyst console and desktop client with React and Tauri, backed by Go (Echo) services, delivering real-time endpoint security visibility across 400+ devices per tenant.

- Contributed to Sentinel's Domain-Driven Design architecture for detecting endpoint vulnerabilities and threats across multi-tenant enterprise networks.

- Designed and built a plugin architecture and SDK for the React monorepo, letting teams extend the security application's core functionality securely.

- Unblocked parallel development and cut feature time-to-market by standardising features and UI components across company applications.

- Centralised content management with Strapi CMS, enabling the marketing team to publish across multiple company websites without engineering involvement.

- Managed static company websites with Hugo and Frontmatter for structured content authoring.

- Developed booking and revenue modules with Next.js, Node.js and Supabase.

- Developed and maintained ERP operations covering appointment scheduling, inventory, HR and finance.

- Built an end-to-end Android POS booking application with React Native / Flutter to streamline in-clinic ordering, booking and billing.

- Engineered a Node.js-backed Zalo Mini App that visualises real-time revenue metrics and customer insights for managers.

- Automated booking and check-in/check-out notifications, plus daily KPI summaries, with a custom Lark bot built on Node.js webhooks.

- Designed and implemented a responsive marketing website and custom CMS for Aura International Beauty Salon using React and Supabase.

- Applied UI architecture best practices to build reusable component libraries and improve overall performance.

- Built the website for Wellbeing Circle VN (wellbeingcirclevn.org), the community-facing project of A Happier Me that extends its social-emotional learning (SEL) resources directly to schools, clubs, and students.

- Supported community engagement and organizational growth for A Happier Me's wellbeing circle model.

- Helped build a Learning Management System (LMS) to deliver digital courses and skills training for students, supporting emotion recognition and healthy interpersonal connection.

- Optimized the user experience for both administrative and learner flows across the platform.

- Led development of an AI-powered visitor-management platform that verifies visitors with real-time facial recognition at check-in and tracks their access.

- Built the frontend with Next.js and React Native, including a mobile-responsive web interface, shipped as an ElectronJS desktop application for real-time visitor check-in and tracking.

- Ensured data accuracy, security compliance and a seamless check-in experience for both visitors and administrators.

- Integrated the Vaidio AI Vision 7.0 facial-recognition engine by developing Java / Spring Boot REST services between the AI engine and the visitor-management application.

- Connected face-match results from the AI engine to the check-in and access-tracking workflow.