Accounting firms are being told to make their work AI-compatible. What that means in practice, why reviewability is the hard part, and a six-stage roadmap showing how staff move from preparing documents to reviewing and advising, with the Document Substrate doing the reading and checking.
How the Document Substrate keeps the AI model away from the numbers, in one page: the model reads and drafts, code calculates and checks, a person signs.
AI product engineer working across strategy, delivery and the people whose work changes, from AI adoption through to production systems in daily use. Available for contract, fractional and advisory roles. Christchurch, New Zealand.
What changed after four months of real client documents: extraction became a separate flow, every extracted figure is traced back to the page it came from, retrieval stopped being a single vector search, and the most useful thing built on top turned out to be a completeness check that no model takes part in.
For accountants and finance people. How a system built on a probabilistic AI model produces figures nobody has to take on trust: the model reads and drafts, code calculates and checks, and every figure on a deliverable comes from a page or from arithmetic over figures that do.
The working principles behind every Orbital system, each with the mechanism that enforces it in the Document Substrate: garbage in, garbage out; doer to reviewer; the model out of the defensible core; personal details never leave the building; ledger, not spreadsheet; one install per customer; small firms first; expertise is captured, not generated.
A collection of short, illustrated guides written for the people who have to approve an AI system rather than build it — how an AI chat actually works, what happens to personal information, how finance teams keep a probabilistic model's numbers exact, and how the Document Substrate puts it all to work.
Why naive document pipelines fail, and how four steps (parse, redact, extract, embed) give you a foundation that lets any LLM work safely on sensitive business documents.
For accounting practices that run on FYI. What FYI does, what the Document Substrate does, the three places they overlap, the one place the AI differs, and what Orbital will not build because FYI already has it.
Applying future-proofing principles to CTI's actual architecture — what to revisit on every model release, where the architecture should evolve, what's missing, and a concrete review cadence. The companion piece that turns the general principle into a working checklist for one production AI system.
What it actually means to 'not build to current model limitations' — the difference between stable constraints worth designing around and temporary weaknesses that age into technical debt. With concrete examples of patterns that survive model upgrades and patterns that don't.
Why 'substrate' is the right word for what an Orbital Map engagement produces: where the word comes from, what it says, and what the alternatives lose.
What 'substrate' means in your field, with worked examples for accounting, law, environmental consulting, logistics, residential construction and wealth management. If you work in one of these fields, use your section as the starting point for your own.
How Orbital's three phases — Map, Architect, Build — were applied to CTI specifically. The case study from the inside: what got mapped, what got architected, what got built, and why the order matters.
How Orbital built an AI coaching platform whose recommendations are grounded in the athlete's own power data, fitness trajectory, and past conversations — not generic advice. The public proof of concept for the approach Orbital takes with every client.
How CTI's Performance Management Chart implementation — Critical Power, Normalized Power, TSS, CTL, ATL, TSB, plus subjective Feeling and RPE — gives the AI coach a quantitative model of fitness, fatigue, and form.
How CTI exposes its rides, fitness state, profile memory, workout generator, and coach over the Model Context Protocol — including the OAuth 2.1 + PAT auth model, the tool/resource/prompt surface, and the security plumbing that keeps it scoped per user.
How CTI's prompt versioning, skill system, admin trace review, Evalite evals, and Langfuse telemetry form a reinforcing loop that makes the AI coaching layer improve with every user interaction.
The case for AI grounded in a business's own judgement rather than a generic model, and the Map, Architect, Build engagement shape that produces it. Originally the Orbital home page, April to September 2026; kept as the long-form version of the argument.
A deep dive into the intent routing, layered prompts, three-tier memory system, and hybrid search that power CTI's AI coaching layer.
Case studies: PII-safe document AI for professional practices, 3D geospatial data visualization, fraud network analysis, production AI pipelines, distributed IoT monitoring, engineering simulation UX. Custom systems for problems off-the-shelf tools could not solve.
Transform indoor cycling training files into an interactive, cinematic experience. Not just charts — a way to discover insights about routes, locations, and performance through exploration.
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.
Practical UX patterns for AI systems that balance autonomy and control. From pure suggestion to observation mode, learn how to design AI interfaces users actually trust—with React examples, confidence indicators, reasoning transparency, and reversible actions.
Why generic AI chatbots fail and how to build bespoke AI applications that solve real problems. Learn the three critical components of successful AI integration - deep system integration, human-in-loop UX patterns, and solving one specific problem extremely well.
Exploring three approaches to LLM state management for TypeScript applications — whole state serialization, persistent memory, or selective context. A Frontend Maximalist perspective on LLM integration architecture and the tradeoffs between simplicity, token costs, and code complexity.
Detailed comparison of Model Context Protocol (MCP) feature support across leading clients — including Claude Desktop, Nanobot, Cline, Cursor, Windsurf, and Postman — examining how each handles tools, resources, prompts, sampling, OAuth, and other key capabilities.
A comprehensive technical guide to building production MCP servers. Learn tools, resources, authentication, MCP-UI, and why MCP is the future of enterprise AI integration.