AI Systems Engineering

Making complex systems explorable through custom AI integration

I'm Johan, a solo engineer in Christchurch, New Zealand, building bespoke AI applications for technical teams whose problems can't be solved with off-the-shelf tools.

Not chatbots. Not demos. Working systems that integrate with your distributed infrastructure and deliver measurable outcomes.


What I Build

Real-Time Visualization for Distributed Systems

Your infrastructure generates data streams that are hard to understand. I build React applications that make complex architectures explorable — combining real-time visualization, intelligent querying, and AI-assisted pattern detection.

Example use cases: Network topology monitoring, IoT sensor dashboards, supply chain visibility, infrastructure observability

Human-in-the-Loop AI Systems

Most AI is either useless chatbots or dangerous automation. I design systems where AI suggests and analyzes, humans decide and refine, and both learn from the interaction.

Example use cases: Engineering design assistants, fraud detection workflows, operational decision support

Project-Scoped Memory and Retrieval

Generic AI forgets everything. I build memory systems that maintain context across interactions, with retrieval pipelines that surface the right information at the right time.

Example use cases: Technical documentation assistants, customer support systems, research platforms

MCP Servers and AI Integration

I create Model Context Protocol servers that make your internal services accessible to AI assistants — databases, APIs, and tools available within conversations, with proper access control.

Example use cases: Internal tooling integration, API accessibility, custom AI capabilities


How We Work Together

I take on 2-3 clients at a time for focused, outcome-driven engagements.

Pilot Project ($25-50K, 6-8 weeks)

Start here. I solve one specific problem to prove value before larger commitments.

What you get:

  • Working prototype solving a defined problem
  • Clear ROI metrics (time saved, costs avoided, revenue enabled)
  • Decision point: extend to full project or walk away

Ideal for: Testing fit, proving concept, immediate high-value problems

Fixed-Fee Project ($80-160K, 3-6 months)

For larger problems requiring complete solutions.

What you get:

  • Production-ready application
  • Documentation and knowledge transfer
  • Defined scope with clear deliverables
  • Measurable business outcomes

Ideal for: Strategic initiatives, new capabilities, system improvements

Ongoing Development ($8-12K/month)

After a successful project, retain ongoing capacity for iteration, new features, and maintenance.

What you get:

  • Priority access to my time
  • Continuous improvement cycles
  • Fast response to changing requirements
  • Strategic technical partnership

Ideal for: Long-term platform development, evolving systems

All pricing in USD


Who This Works For

You're a Good Fit If:

  • You have a specific problem that off-the-shelf AI tools can't solve
  • You work with complex systems (distributed architectures, real-time data, geospatial, graphs)
  • You value outcomes over process — you want things shipped, not endless meetings
  • You have budget for specialized work ($25K+ minimum engagement)
  • You're technical enough to articulate the problem and evaluate solutions

Target Sectors:

Environmental monitoring and geospatial · Engineering design tools · Industrial IoT and sensors · Infrastructure and observability · Fraud detection and financial monitoring · Supply chain and logistics · Scientific computing and research

You're NOT a Good Fit If:

  • You need generic chatbots or "AI wrappers"
  • Your problem can be solved with existing SaaS tools
  • You want staff augmentation or hourly contractors
  • You're looking for generic web development
  • You need someone to attend daily standups

If existing tools would work, I'll tell you that.


Technical Approach

Core stack: React/TypeScript · Real-time data handling · LLM orchestration and prompt engineering · Distributed systems architecture · Geospatial and network visualization · MCP server development

Philosophy: Ship products first, build infrastructure later. I focus on solving your specific problem with working code, not building elaborate frameworks or perfect architectures.

Working style: Remote-first from Christchurch, New Zealand (APAC timezone friendly). Direct communication, no corporate consulting theater. You get me, not a team of juniors.


Recent Work

Building with MCP How to connect AI assistants to real infrastructure using Model Context Protocol servers.

Frontend Maximalism and LLM Context Architectural patterns for integrating LLMs into React applications without losing control.

Example Projects Custom AI applications and React engineering work demonstrating the approach.


Why Not Just Hire Someone?

You're probably here because:

  • You've been trying to hire a senior AI engineer for 3-6 months
  • You've talked to agencies who want to sell you generic chatbots
  • Your team doesn't have the specialized expertise for this specific problem
  • You need something working now, not after a long recruitment process

Here's what working with me gets you:

  • Speed: 6-8 week pilot vs 4-6 month hiring cycle
  • Clarity: Fixed price for defined outcomes vs uncertain salary + time commitment
  • Specialization: Rare combination of React + AI + distributed systems expertise
  • Risk mitigation: Pilot project proves value before larger commitment

The math: A failed 6-month hire costs you ~$100K in salary + opportunity cost of delayed outcomes. A $25-50K pilot either solves your problem or tells you definitively what won't work.


Location & Availability

Based in Christchurch, New Zealand APAC timezone (UTC+12/+13) with good overlap for Australian and Asian clients. Also work with US/European remote-first companies on async schedules.

Availability: Currently taking on 1-2 new pilot projects. Pilot projects can typically start within 2-3 weeks.


Let's Talk

If you have a specific problem that might be a fit, email me with:

  1. What specific problem you're trying to solve (not just "we want AI")
  2. Why off-the-shelf tools haven't worked
  3. What outcome would make this worth $25-50K to you

I respond to every serious inquiry within 24 hours.

Email: info@orbital.co.nz


Common Questions

Can't we just use ChatGPT/Claude/other existing tools? Maybe! If existing tools would work, I'll tell you that. I work on problems where off-the-shelf doesn't cut it — usually because you need deep integration with your specific systems, human-in-loop UX patterns, or specialized retrieval/memory.

Why these price points? Bespoke AI integration requires rare expertise (React + AI + distributed systems). You're paying for specialized knowledge that saves months of trial-and-error and delivers systems that actually work. The alternative is hiring (4-6 months, $100K+ cost) or agencies (generic solutions, longer timelines).

Do you work hourly? No. Fixed-fee projects or monthly retainers only. You get predictable costs and I focus on outcomes instead of billing time.

Can you do a smaller project first? Yes — that's what pilot projects are for. $25-50K to solve one specific problem and prove value before larger commitments.

What if our budget is smaller? If you have less than $25K, I'm probably not the right fit. Consider: hiring a mid-level developer, using existing SaaS tools, or waiting until the problem is expensive enough to justify specialized help.

Are you available for ongoing work? After a successful pilot or project, yes. I work with 2-3 clients on ongoing development retainers. This ensures you have priority access when you need new features or iterations.

Do you work with startups? Sometimes. If you're pre-revenue or early-stage, the minimum engagement is usually too high unless you've raised funding specifically for this problem. Better fit: you're post-PMF with engineering budget and a specific integration challenge.


Bottom Line

Generic AI tools aren't solving your problem. You need something built specifically for your systems, your data, your workflows.

I build exactly that — one project at a time, with clear outcomes and measurable ROI.

Email: info@orbital.co.nz


Latest Articles

Building AI Products That Ship: From Insight to Production in Weeks

Ship useful AI products in weeks, not months. Learn the agent lab architecture, how to identify net-new work opportunities, make your data AI-ready, and execute a 4-8 week development cycle from insight to production. Focuses on outcomes over capabilities, deep integration over generic tools, and rapid iteration with real users.