AI for Practice Staff

This essay sits primarily under the How do we do AI? question of the practice’s methodology: practical patterns for the practice staff whose work AI most directly intersects.

Quick start

If you are short on time, start here.

1. Choose your situation:
   • I have data to make sense of → use the Data Analysis pattern below.
   • I am reviewing or writing policy → use the Compliance & Policy pattern.
   • I need to communicate a change → use the Stakeholder Communication pattern.
   • I am planning a change to a process → use the Strategic Analysis or Process Improvement patterns.
2. Use the relevant template, paste a short description of your context, and fill in the brackets.
3. Ask AI for a first draft, then:
   • Check for accuracy and alignment with the firm’s policies and obligations.
   • Adjust tone, detail, and length for the people who will read it.
4. Save anything that works so you can reuse it.

A starting prompt:

Context: [Brief description of the task and audience]
Role: You are a [my role: practice manager, paralegal, etc.]
Task: Help me [analyse / summarise / draft] [the thing]
Output: [One-page summary / email / checklist] in clear, practical language

Why this guide is for you

If you are a practice manager, paralegal, legal secretary, bookkeeper, practice nurse, receptionist, client coordinator, or any of the roles that hold a small regulated firm together, this is for you.

While much of what people say about AI focuses on practitioners (how solicitors, clinicians, accountants, and advisers use AI on the work itself), your role is different. You work across the practice and often manage:

• Data analysis and reporting
• Compliance and policy coordination
• Stakeholder communication
• Process improvement and operations
• Documentation and templates

AI can support these responsibilities by amplifying your expertise.

What makes practice staff roles different

Your value is that you see patterns across the firm that individual practitioners do not. You understand the practice management system, the rhythm of the work, what is breaking, and how stakeholders actually behave. You translate between practitioners and the rest of the world. You solve the problems that affect scale.

AI helps you do all of that better and faster.

Core AI tasks for practice staff

1. Data analysis and insight

You might:

• Analyse matter or appointment trends
• Review WIP, ageing, utilisation, recovery rate
• Identify at-risk matters or patient-flow bottlenecks
• Compare metrics across practitioners or service lines

How AI helps:

• Summarise data quickly
• Identify patterns
• Surface outliers
• Generate clear summaries for partners
• Suggest the next question to ask

A simple workflow:

Step 1: "Summarise this quarter's WIP and matter-stage data
(attached CSV). Highlight matters at risk of slipping."

Step 2: "Now identify any patterns: practitioner, matter type,
age. What is going wrong?"

Step 3: "For each pattern, suggest two practical interventions
inside our existing capability."

Step 4: "Turn this into a one-page partner-meeting summary
with clear recommendations."

(See Prompt Chaining for the full pattern.)

2. Compliance and policy

You might:

• Track regulatory obligations (LPUL, AHPRA, APES 110, or RG 175, depending on sector)
• Assess current practice against standards
• Identify compliance gaps
• Draft or update internal policies
• Translate regulator guidance for practitioners

How AI helps:

• Summarise complex policy documents
• Identify key requirements
• Flag risks and gaps
• Draft policy updates and internal guidance
• Translate technical regulation into plain English

Step 1: "Paste of [regulator guidance]. Extract the five most
critical requirements for our firm."

Step 2: "Compare those against our current internal policy
(attached). Where are the gaps?"

Step 3: "For each gap, suggest a specific update with
implementation timeline."

Step 4: "Draft updated guidance for practitioners. Plain
language. They are busy."

3. Stakeholder communication

You might:

• Draft internal emails, newsletters, partner-meeting agendas
• Communicate process changes to practitioners
• Create client-facing guidance
• Prepare board or partner-meeting summaries

How AI helps:

• Draft communications and adapt tone for the audience
• Reformat technical content for different readers
• Create multiple versions (executive summary, full email, client-facing handout)
• Generate talking points for difficult conversations

Step 1: "I have this compliance update (quite technical).
Draft a two-paragraph executive summary for partners."

Step 2: "Now create a one-page email for practitioners
explaining what changed and what they need to do."

Step 3: "Create a brief note for client-facing staff so
they can answer questions consistently."

Step 4: "What three concerns should I expect from
practitioners and how would I address each?"

4. Strategic planning and operations

You might:

• Analyse resource allocation
• Plan a new service line or staffing change
• Design or update workflows
• Evaluate the effectiveness of a recent change

How AI helps:

• Compare options side-by-side
• Generate implementation plans
• Identify risks and mitigations
• Create evaluation frameworks

Step 1: "We are considering moving to four-day weeks for support
staff. Generate three implementation options."

Step 2: "For each, what would implementation require?
Timeline, cost, partner buy-in?"

Step 3: "What risks does each option carry?
How would we mitigate them?"

Step 4: "Create a one-page comparison for the partner meeting."

5. Documentation and process improvement

You might:

• Document existing processes
• Design new workflows
• Create guidance, templates, or checklists
• Improve clarity of existing procedures

How AI helps:

• Draft procedural documentation
• Identify gaps
• Suggest improvements
• Generate clear, accessible internal guidance

Step 1: "Here is our current intake process. Map what actually
happens step by step."

Step 2: "Where are the bottlenecks? Where do things stall?"

Step 3: "Suggest a streamlined process with clear decision
points and hand-offs."

Step 4: "Create a one-page checklist staff can use during
intake."

Five practical patterns

Pattern 1: data analysis

Context: [Dataset: what, when, why collected]
Role: You are a [practice manager / coordinator / analyst]
Task: Analyse [specific data]. Focus on [the questions].
Output: [Summary / table / bullets]
Audience: [Partners / practitioners / external]

[Paste data]

First step: [summarise / identify trends / flag issues]

Pattern 2: compliance and policy

Context: [Regulatory or policy document]
Role: You are a [compliance / policy / quality lead]
Task: Extract key requirements; assess current practice; flag gaps
Output: [Summary / checklist / gap analysis / action plan]

[Paste document or describe requirement]

First step: identify the requirements

Pattern 3: stakeholder communication (multi-format)

Content: [The information to communicate]
Task: Create three versions:
1. [Audience A]: [format] with [tone/focus]
2. [Audience B]: [format] with [tone/focus]
3. [Audience C]: [format] with [tone/focus]

Original content: [paste]

Start with version 1.

Pattern 4: strategic analysis

Scenario: [Constraint or opportunity]
Role: You are a [practice operations manager / coordinator]
Task: Generate [n] options for [goal].
For each: [cost, risk, timeline, effectiveness, partner buy-in]
Recommendation: [your judgement]

Scenario detail: [describe situation, constraints, objectives]

Pattern 5: process documentation

Current process: [brief description]
Problem / goal: [what's not working / what to improve]
Role: You are a [practice operations specialist]
Task: Map current state; identify issues; redesign
Output: [Flowchart / checklist / step-by-step guide]

Details: [describe current process and pain points]

First step: map the current process step by step

Real workflow: end-of-quarter analysis

Scenario: End of quarter. The senior partner wants a quick read on practice performance and emerging risks.

Step 1: initial analysis

Attached is the matter-management export for Q1. Summarise:
- Practitioner utilisation (highest, lowest, anomalies)
- Matter ageing: anything stalled longer than usual?
- Realisation rate: overall and by practitioner

Plain bullets. What does the partner most need to know?

Step 2: deeper read

You flagged three matters as significantly stalled. For each, what
factors might cause this, looking at the data alone? What follow-up
should I do before the partner meeting?

Step 3: partner-meeting summary

Turn this into a one-page summary for the partner meeting. Include:
- The headline number
- Three patterns worth noticing
- Two practical recommendations
- The questions I am still working on

Keep it readable in two minutes.

Step 4: the partner’s likely questions

What are the three questions the senior partner is most likely to
ask about this summary? Help me prepare a one-line answer for each.

That sequence turns several hours of preparation into something usable in under an hour, with the staff member’s professional judgement applied at each step.

Principles for practice staff using AI

1. Use AI for analysis and insight, not decisions. AI can help you understand data and generate options. Partners (and you, in your role) make the decisions. Your value is knowing the firm; AI’s role is processing and drafting.

2. Maintain professional judgement. Do not accept AI output at face value. Verify facts, especially anything regulator- or obligation-related. Consider firm-specific context. Combine AI output with what you actually know.

3. Use chaining for complex tasks. Compliance reviews, strategic planning, multi-stakeholder communication: break them into steps. Quality compounds.

4. Use a prompt framework for individual prompts. CRAFT or RTCF: pick one and stick to it.

5. Protect sensitive information. If working with personal data, client matter information, patient records, or anything covered by Privacy Act, professional obligations, or court order:

• Do not paste actual identifying details into consumer AI tools
• Use the firm’s enterprise tier
• Consider the firm’s AI use policy (if you do not have one, that is the conversation worth having)

6. Document what works. Keep your best prompts. You will use them again.

7. Combine AI with your expertise. You know how the firm actually works. You know what partners actually care about. You know what is feasible given the constraints. AI does not. Your judgement applied to AI’s output is the real product, and your willingness to steer the work, not just refine the model’s first draft, is what keeps you in the role partners actually rely on.

What not to use AI for

• Confidential personnel decisions without appropriate safeguards
• Final decisions that require judgement: AI informs; you (or the partner) decide
• Anything client- or patient-affecting without practitioner review
• Legal interpretation beyond explaining what a rule says: get the lawyer
• Anything regulated that you cannot defend in your own words afterwards

Getting started

Choose your starting point:

• I have a data question → Pattern 1
• I am reviewing a policy → Pattern 2
• I need to communicate something → Pattern 3
• I am planning a change → Pattern 4
• I am documenting a process → Pattern 5

Then: follow the template; review the worked example above or in the linked guides; adapt to your firm; try it with a current real task.

Final thought

Your role is to enable good professional work at scale. AI helps you see patterns, analyse complex data, draft communications, and explore options faster than was possible. But you are the expert on the firm, its people, and what is realistic. Use AI as the analytical partner, not as a substitute for professional judgement.

The most effective practice staff combine AI’s analytical power with deep firm knowledge.

About this work

borck.consulting helps regulated Australian small firms develop AI capability across the whole practice (partners and staff) so the firm gets the benefit, not just the latest pilot. Engagements begin with the AI Readiness Diagnostic.