CUSTOM BUILD
A custom app, AI agent, or connected workflow built around your business.
You need more than a single improvement: an application or workflow that brings several moving parts together.
A build can combine several data sources with analysis and recommended actions, as in the Entec hub. We agree scope and timing before work starts.
WHAT YOU’RE INVESTING IN
The most valuable part of a custom system often isn’t the AI. It’s getting the rules out of people’s heads and into a form the business can use, connecting information across your systems, and agreeing what matters, how it’s calculated, and what should happen next.
Most of that foundation is conventional software: data connections, calculations, business rules, and workflows. AI handles the parts that benefit from interpretation, such as extracting information, assessing relevance, or summarizing correspondence.
That separation matters. As models improve or pricing changes, we can evaluate and update the AI component without necessarily rebuilding the business logic and connections around it. Switching still needs testing, and some changes need development, but the knowledge and structure don’t have to be recreated because a new model arrived.
You’re investing in a system that captures how your business works, with AI as one adaptable part of it.
CONVENTIONAL SOFTWARE
AI, WITHIN AGREED LIMITS
These aren’t two separate systems. The foundation gives the AI its context, and conventional code decides what happens with what the AI produces. Both need testing: ordinary code is only as right as the rules we agreed.
In the Entec hub, a better model might improve how the system interprets an inquiry or summarizes a customer conversation. It wouldn’t replace the work of connecting SimPRO, Xero, and Outlook, defining key accounts, or agreeing which follow-ups deserve attention.
CUSTOM AI AGENTS
An agent could research prospective customers and prepare account briefs, or monitor incoming inquiries and draft the next step. I design the workflow around your information and tools, with clear boundaries on what it can do and when it needs your approval.
For example, a customer-research agent could check sources you approve, compare prospects against your criteria, and prepare a brief explaining why an account might be worth a conversation. You review the evidence before deciding who to contact. This is an example of a possible build, not a client case study.
Before launch, we test representative tasks and exceptions, including missing or conflicting information. Ordinary code handles exact calculations and fixed rules; AI handles interpretation where it helps. We agree access limits and approval steps, and keep a record of the agent’s actions so you can check what happened.
The point is to give a small team research coverage it couldn’t previously justify staffing, without handing over its commercial judgment.
GROUNDED IN ACTUAL WORK
built around the business.The problem. Customer activity sat across SimPRO, Xero, and Outlook, with no single view of which accounts and inquiries needed attention.
What changed. A working hub that tracks key customers and inquiries, and prioritizes the next commercial action.
Read the story
The problem. An analyst needs to compare more than 40 metros across property sectors and explain what is driving each one.
What changed. Comparable scores and a written read of the trends worth investigating.
LET’S TALK
You don’t need a technical brief. Tell me what you’re trying to do, and we’ll work out a sensible next step.
