Hi, I'm Lan Nguyen Si

// Software Architect

Building production-grade Human-Agent Operating Systems

10+ years in software engineering, including 3 years of architecture expertise in regulated GovTech - now applied to trustworthy, self-hosted agentic AI infrastructure.

StatusAvailableMagdeburg, DEFocusAgentic AI Infrastructure
01About

About Me

Portrait of Lan Nguyen Si

Lan Nguyen Si — Magdeburg, DE

Software engineer with 10+ years of experience designing scalable, production-grade systems, including 3 years as a software architect in regulated GovTech. In parallel, I develop Project OS - an open-source Human-Agent Operating System that brings reliable task orchestration, agent grounding, multi-agent collaboration and self-hosted deployment to life. I combine architectural expertise from large-scale government platforms with deep hands-on work in agentic AI to create trustworthy human-AI solutions.

Current Role
Software Architect at publicplan GmbH - building national digital government platforms.
Side Initiative
Creator of Project OS - an open-source Human-Agent Operating System.
Location
Based in Magdeburg, Germany. Working remotely with teams across Europe.
Education
B.Sc. Computer Science from Otto-von-Guericke-Universität Magdeburg.
02Experience

Experience

My professional journey in software development

Jul 2022 - Present
Current

Software Architect / Senior PHP Engineer

publicplan GmbH · Remote (Saxony-Anhalt) · Full-time

Acting as Software Architect for key components of Germany's digital government infrastructure. Designing and implementing scalable, reusable SaaS components for the national Register Modernization and the Business Service Portal NRW (Once-Only Technical System / OZG). Establishing architectural patterns and technical standards for distributed, containerized systems serving federal and EU-level use cases. Parallel initiative: Developing Project OS to explore next-generation architectural patterns for human-AI collaboration in complex enterprise systems.

PHPSymfonyDockerKubernetesMariaDBMongoDBReactNext.jsNode.jsRedis
Jan 2018 - Jun 2022

Software Engineer

IServ · Hybrid · Full-time

Core development of the IServ3 platform, scaling support from 2,000 to 5,000 schools. Developed critical modules (Email, Calendar, CRM) and optimized backend logic. Mentored team members through code reviews and technical in-house talks.

PHPSymfonyJavaScriptPostgreSQLMySQLLinuxjQueryVue.js
Mar 2016 - Dec 2017

Software Engineer

Invia Travel Germany GmbH · Magdeburg · Full-time

Promoted from Junior to Software Engineer in Jan 2017. Lead developer for 'Travellogins' and 'German Travellersclub Card' platform. Implemented high-performance backend functionality and API integrations for major travel portals.

PHPZend FrameworkREST APIsSOAPJavaScriptMySQLCache
03Stack

Skills & Technologies

Technologies and tools I use to build modern, agentic AI and enterprise systems

Agentic AI & Orchestration

Multi-Agent SystemsTask OrchestrationAgent GroundingHuman-in-the-LoopConfidence ScoringMemory ManagementTool UseMCP Bridge

Platforms & Frameworks

Project OSNext.js 15HonoPrismaTypeScriptNode.jsReact

Infrastructure & Production

DockerKubernetesTraefikRedisPostgreSQLClickHouseCI/CDOIDCLinux

Backend & Legacy

PHPSymfonyREST APIsGraphQLMariaDBMongoDBAgile/Scrum

Certifications

  • Certified Professional for Software Architecture (iSAQB)
  • ITIL 4 Foundation Certificate in IT Service Management
  • Certified PHP Developer
04Work

Featured Projects

From open-source agentic AI infrastructure to production government platforms.

Building Now

01

OpenTriologue

Active·2026
AI Agents

Real-time platform where humans and AI agents collaborate as real teams: shared rooms, Bring-Your-Own-Agent (BYOA) support and semantic memory.

  • Bring-Your-Own-Agent (BYOA)
  • Shared human-AI workspaces
  • Semantic memory
Next.js 15TypeScriptHonoPrismaPostgreSQLDocker
02

Project OS (agent-tasks)

Active·2026
AI Agents

Open-source Human-Agent Operating System: declarative workflows, claim-gates, confidence scoring, GitHub integration and an MCP bridge for reliable task orchestration.

  • Declarative agent workflows
  • Confidence scoring and claim gates
  • MCP bridge
Next.js 15TypeScriptHonoPrismaDockerMCP
03

Agent Harness Control Plane (harness)

Active·2026
AI Agents

Declarative control plane for agent harnesses: one YAML for grounding, tools, memory and hooks. Describe, validate, diff, apply.

  • Declarative agent config
  • Validate, diff, apply
  • Grounding and policy as code
TypeScriptNode.jsYAML
04

agent-grounding

Active·2026
AI Agents

Verification framework for AI agents: runtime checks, claim validation and hypothesis tracking to reduce hallucinations in autonomous agents.

  • Runtime verification
  • Claim validation
  • Hypothesis tracking
TypeScriptNode.jsMCP

Production Track Record

05

Business Service Portal NRW

In Production·2024
Government Tech

Central digital platform for business services in North Rhine-Westphalia (WSP.NRW). Implementing the Online Access Act (OZG) to digitalize administrative processes for companies.

  • Official OZG Implementation
  • Microservices Architecture
  • High Security Standards
PHPSymfonyDockerKubernetesElasticsearchReactNext.jsNode.js
06

Register Modernization (RegMo)

In Production·2023
Government Tech

Implementation of the national Register Modernization project. Developing components for the 'Once-Only' technical system to enable data exchange between authorities.

  • EU Once-Only Principle
  • Cross-Border Interoperability
  • Modern API Standards
PHPSymfonyJavaDockerKubernetesOpenAPIRedisMariaDBMongoDBReactNext.jsNode.js
07

IServ School Platform

Completed·2021
EdTech

Core development of the IServ3 school server platform. Responsible for critical modules including the Email system, Calendar, and Customer Management.

  • Used by 5000+ Schools
  • High-Availability Email System
  • Legacy Migration
PHPSymfonyJavaScriptMySQLLinuxDebianBash

Interested in Agentic AI, Human-Agent Systems or production-grade AI infrastructure? Let's talk.

Let's Discuss Your Project
05Contact

Let's build something reliable.

Interested in Agentic AI, Human-Agent Systems or production-grade AI infrastructure? Let's talk.

Local time--:-- CET