What it is

AI Copilot is an optional assistant that helps you draft and refine content directly inside Rico’s text fields. It uses a large language model (LLM) from OpenAI and is designed to work with Rico’s formatting and templating so your output stays consistent with your report style.

How it works

• Available wherever you type in Rico’s text fields.

• Understands limited page context (e.g., the current workflow step/instructions) to produce better-targeted drafts.

• Today’s model is a general LLM (not trained on your projects). We’ll make it smarter over time.

Why use it 

• Compatible with Rico’s formatting, headings, bullets, and templates.

• Speeds up first drafts and routine edits.

• Keeps you in the driver’s seat- everything is editable, and your professional judgement remains essential.

Identified use cases (evolving)

You’re free to experiment, but common patterns include:

• Transform & tidy: “Summarise into a table”, “Clarify this paragraph”, “Condense to 150 words”, “Turn notes into bullet points.”

• Draft a template from instructions: “Create a template for this step’s instructions with headings, bullet lists, and fill-in-the-blank fields.”

• Populate or adapt a template: “Adapt this template for a proposal with earthworks, infringing vehicle accessway, demolition of the existing dwelling, and construction of four new dwellings.”

• Generalise an existing response: “Turn this past response into a reusable template for new projects.”

Prompt tips

• Be specific about the goal (“two paragraphs, plain language, include assumptions”).

• Provide the facts you want included; don’t ask the model to infer planning rules you haven’t supplied.

• Ask for structure (“H1/H2 headings, bullets, and a short conclusion”).

• Iterate: if the first result isn’t right, say what to change (“shorter”, “NZ English”, “keep policy references verbatim”).

Things to watch out for

• Technical accuracy matters. Planning assessments are precise; the copilot can draft, but can make mistakes, omit nuances, include false facts and make incorrect assertions.

• You must review outputs. Treat results as a starting point, not a final decision.

• Clear inputs win. The clearer your instructions and templates, the better the results.

• Don’t over-delegate. Use the copilot for structure, clarity, and speed - keep determinations and citations under your control.

Data handling and privacy (summary)

• Prompts and outputs are processed by OpenAI solely to generate your result.

• Your content is not used to train their models.

• Your data remains within your organisation’s Rico workspace (not shared with other Rico customers)

• AI functionality is optional. We’ve done our best to put in the right protections and guardrails around your data, but we’re also committed to giving you choice. AI functionality can be turned off or on for your organisation at your request.

If you’d like more information or have any questions, don’t hesitate to get in touch.

Getting started (opt-in)

1. Let us know you’d like to opt-in and we’ll turn the feature on for your organisation.
2. Pick low-risk sections first (boilerplate, clarifications, summaries).
3. Use the prompt tips above and your existing step instructions/templates.
4. Review and edit every output before use.

Roadmap (what’s next)

• Feedback: we’ve left space in our roadmap to respond to your feedback, so please let us know how you’re getting on and what you’d like to see more / less of.

• Better use Rico data: we’re looking to make it easier for you to use Site, Plan Assessment and Previous Project Data in prompts to improve AI results.