Orbital
TLDRArticlesAbout

Johan Steenkamp

Summary

I build React applications and AI systems that make complex systems explorable and understandable, with a focus on geospatial data fusion.

With over a decade of experience, I transform complex distributed systems into interactive experiences through React visualization, geospatial data fusion, MCP integration, and production AI systems. I've led engineering teams, shaped technical practices for modern React and AI development, and delivered business-critical solutions across fraud detection, industrial monitoring, and engineering design tools.

I focus on three areas: building React applications with strong patterns and performance, developing MCP servers that make services accessible to AI, and creating outcome-driven AI systems that solve specific problems, including fusing disparate intelligence sources into unified, explorable spatial views.

Principles

Development Principles

  • Geospatial + Agentic: The intersection is underserved. Traditional GIS is slow and siloed; general AI lacks domain depth. Geospatial data fusion, overlaying signals, imagery, network data, and entity locations onto a single map, is where the real value lives.
  • Make data explorable, not just displayable: Fusion is not just about seeing data; it's about reducing cognitive load. Show the most important information first, enable deeper dives, and surface temporal and network context alongside spatial context.
  • Build systems agents can navigate: Every implementation should be AI-accessible, self-documenting, self-maintaining. The future of intelligence fusion is AI doing the cognitive work, with humans in the loop for validation.

T-Shaped IC Principles

  • One person who ships beats a coordinated team: The builder archetype matters more than ever. Own the full stack from problem to solution.
  • Custom workflows are portable assets: Your agentic patterns travel with you, making you 5-10x more effective anywhere.

Geospatial Data Fusion

Geospatial data fusion is the ability to take disparate types of intelligence, signals, imagery, population data, network graphs, and entity locations, and overlay them onto a single map for a unified, actionable view. This is not just cartography; it is the core challenge of making complex real-world systems understandable.

The problems I focus on:

  • Reducing cognitive load: Analysts and operators are overwhelmed by data. Fusion means surfacing the right information at the right time, not dumping everything onto a screen.
  • Temporal and network context: Space alone is not enough. Understanding when events happened and who is connected to whom is what separates intelligence from data.
  • Gray zone activity: The most interesting use cases involve subtle, non-obvious patterns: commercial vessels behaving like adversaries, infrastructure anomalies, supply chain disruptions. These require fusing multiple data streams that no single source captures.
  • AI-native fusion: The next step is not just visual overlay but cognitive fusion, where AI synthesizes across data sources and surfaces conclusions, while keeping humans in the loop for decisions that matter.

Traditional GIS tools were not built for this. Custom React applications, backed by AI systems with proper context and memory, are.

Experience

AI Application Engineer

Orbital | Self-employed Apr 2025 – Present

  • Building production-ready React applications with real-time data visualization, including geospatial data fusion interfaces, network graphs, and multi-source intelligence overlays
  • Developing MCP servers that connect AI agents to existing services and data sources
  • Creating outcome-driven AI systems with evals, observability, and prompt management for reliable production deployment
  • Specializing in making distributed systems explorable through interactive visualization and intelligent tooling, reducing cognitive load by fusing disparate data into unified spatial views
  • Rapid prototyping and PoC development to validate ideas and accelerate time-to-value
  • TypeScript, React, Next.js, AI SDK, AI Gateway, LLM APIs, MCP, Tailwind CSS, shadcn/ui

Principal Frontend Engineer

Darwinium | Full-time Sep 2022 – Apr 2025

  • Led development of the fraud detection frontend, UI, dashboards, and data visualization
  • Shaped engineering practices for modern React state management, UI and data visualization libraries, and CSS modules
  • Collaborated with cross-functional teams to deliver business-critical solutions
  • TypeScript, React, GraphQL, AntD, AntV, Mapbox

Digital Architect & Head of Software

Syft Technologies | Full-time May 2018 – Sep 2022

  • Responsible for the end-to-end process of driving, building, and improving Syft's Digital Infrastructure and Digital Strategy
  • Led the development and implementation of the software strategy, building and growing strong teams
  • Architected and re-architected monolithic products into layered architectures using modern application approaches (cloud, services, messaging)
  • Established core Data and Instrument platform teams and delivered a production-ready Data Platform on AWS
  • Active developer in full-stack applications using React, GraphQL, AWS Amplify, and Next.js

Principal Engineer

Wynyard, Cognevo, & Telstra | Full-time Aug 2014 – Apr 2018

  • Served as Principal Engineer and Application Architect for multiple technology companies, focusing on frontend and UI architecture
  • Led the development of data visualization frontends for security products using a modern stack, including React, Redux, and GraphQL
  • Developed reusable UI components and designed application APIs to allow end-user configuration of the frontend
  • Actively drove the adoption of GraphQL as a core part of the tech stack
  • Provided technical expertise and consulting services to internal product development teams, including code reviews and architecture design
  • Node.js, React, Redux, GraphQL, Bulma CSS

Skills

  • Frontend: React, Next.js, TypeScript, Tailwind CSS, shadcn/ui, GraphQL, Mapbox, Deck.gl, AntD, AntV (G2, G6)
  • Geospatial: Multi-source data fusion, temporal-spatial analysis, network graph overlays, interactive map intelligence
  • Backend: Node.js, AWS, DynamoDB, S3, Postgres, SQLite, Supabase (database, auth, storage), Clerk
  • AI: AI SDK, LLM APIs, Model Context Protocol (MCP), Prompt/Context Engineering, SKILLs, Image Generation

Education

BSc Eng (Electronics) University of Cape Town

LinkedIn

LinkedIn Profile

Example Projects

Examples of the applications I've built:

  • Routebook: Cinematic satellite visualization and fitness tracking for Rouvy cyclists, with fly-along route animation, performance overlays, POI creation, and automatic performance metrics
  • Deep Search: Production-grade search app with evals, observability, and prompt management
  • Darwinium Frontend: Fraud detection UI with interactive dashboards and network graphs
  • Syft Instrument Health: Mass Spectrometry instrument performance monitoring
  • Mersen R-TOOLS MAXX: Heat sink design and thermal simulation tool

Example Application Descriptions and Screenshots

About