How custom AI Agents automate the boring (and unlock the valuable)
Many New Zealand businesses are under pressure to maximise effectiveness, whilst keeping a tight reign on costs.
In this respect AI has been a bright spot, however the majority of businesses haven’t yet gone beyond just their people using various AI tools in an ad-hoc way.
That has probably sped up some operations, but AI agents offer an opportunity to fully automate workflows that were once completely beyond what software could achieve.
The potential exists to free capable people from repetitive, low-value tasks that divert from higher value activities and sap morale.
This is why there’s so much excitement about AI agents revolutionising workflows; it’s definitely the future of productivity but, for forward thinking businesses, the opportunity is here right now.
What Are Custom AI Agents?
You can think of custom agents as little cloud apps that bring AI capabilities together with the power of conventional software and data integrations.
Agents can automate workflows that include the use of AI models to carry out steps of the process, and potentially also to determine what to do next within a decision tree.
It’s easiest to explain by giving a simple example.
Imagine a business onboarding new customers, where each customer starts by filling out an online questionnaire.
THE CLIENT
- Fills in the questionnaire.
A STAFF MEMBER
- Downloads the results from the questionnaire software
- Copies them into an AI, such as ChatGPT, using a pre-made prompt to consistently explain to the AI what to analyse and report back on
- Gets a report back
- Copies that report into a document
- Emails the document to the customer and ask them to choose a meeting time to review (via a scheduling tool)
No big deal when there’s only a few, but there’s actually quite a few steps so this will get repetitive and time consuming if the number of new customers to onboard grows.
An AI agent could automate the whole process.
THE CLIENT
- Fills in the questionnaire.
AN AI AGENT
- Picks up output from the questionnaire software
- Sends it to the AI, along with pre-set prompts
- Receives the report back from the AI
- Formats it into a document and stores it
- Emails it to the customer with a cover note, including link to scheduling tool
This example would be a fully autonomous agent, however let’s say a human needed to review and edit the report before it went out. If this was the case the AI could involve a human in that step of the workflow.
So you can probably envisage how an AI agent automates repetitive tasks, that include one or more AI models in the process, and could also include human input. And, because they’re software, there is also scope to include whatever decision making logic might be needed in the process.
Why mid-sized companies are poised to win
Large corporations have big IT budgets and have been building automation for decades. Small businesses often have low scale and are therefore well served by off-the-shelf AI tools. But mid-market companies often sit in the “messy middle”; complex enough to have real process pain, but not always well-served by standard solutions.
This is where custom AI agents and integrations have great potential to revolutionize productivity:
- Systems are typically already in place (ERP, CRM, accounting, operational platforms).
- Processes tend to be unique; often related to a competitive advantage.
- Teams need automation that fits their way of working, not the other way around.
Custom AI agents bridge this gap, working seamlessly with existing tech stacks and company processes; even potentially learning elements of company culture.
They are also lower cost and higher ROI than a lot of traditional software development because they are generally lightweight from a coding perspective, yet leverage the vast power of AI.
The ROI of Automating the Boring
Here’s what New Zealand companies implementing AI agents are finding:
- Time saved: Often hundreds, or even thousands, of staff hours a year.
- Error reduction: Fewer manual touchpoints mean far fewer mistakes.
- Faster decisions: Data gets into leaders’ hands faster, allowing faster responses to changing conditions.
- Talent retention: Staff are happier when they spend time on meaningful work, not digital drudgery.
For a business with eight-figure revenue, even efficiency gains of 5-10% can translate into hundreds of thousands of dollars a year. Or, in highly competitive markets, a company may choose to add less to its bottom line, focusing instead on shoring up its ability to compete and grow market share.
Some Practical Examples
Customer Engagement & Sales
- 24/7 Sales Assistants: Agents trained on a company’s product catalog, pricing models, and sales scripts that qualify leads, answer detailed product questions, and even create quotes. This can include website chatbots and/or call centre voice-based agents (which are actually easier to communicate with than many offshoring options).
- Hyper-personalized Marketing: Agents that segment customers and generate campaigns based on CRM data, purchase history, and browsing behavior.
- Staff onboarding: Guiding new staff through internal processes, policies and systems, all in a way that is native to your company culture.
Operations & Efficiency
- Internal Helpdesk Agents: Employees can ask “How do I file a leave request?” or “Where’s the latest compliance form?” and get instant, correct answers from an internal knowledge base.
- Automated Process Coordinators: Agents that handle workflows across your software suite, including nudging humans or triggering system actions when things stall.
- Meeting & Project assistants: Custom AI that extracts actionable decisions, deadlines, and risks from meeting transcripts tailored to the company’s terminology.
Finance & Compliance
- Custom Risk Assessment Agents: Monitoring transactions and flagging anomalies based on industry specific thresholds and compliance rules.
- Policy Adherence Agents: Checking contracts, emails, or proposals against company policies or regulatory standards (e.g., NZ privacy law or industry codes).
- Forecasting & Reporting: Pulling financial data from multiple systems and preparing management-ready dashboards or board packs.
Industry-Specific Examples
- Construction: Site safety AI agent that pulls from regulations and company safety manuals to guide people in real-time.
- Healthcare: Patient intake triage agent that aligns with a clinic’s exact protocols while protecting sensitive data.
- Legal: Contract review agent trained on a firm’s templates, highlighting risks in plain English.
- Manufacturing: Predictive maintenance agent monitoring IoT sensor data, alerting only when thresholds specific to the exact machinery are crossed.
Strategic & Leadership Support
- Executive Briefing Agents: Summarize market news, competitor filings, or government updates specifically for a given sector.
- Scenario Planning: Model “what if” analyses based on a company’s historical data, not generic economic models.
- Board Paper Drafting: Pull the right insights and format them exactly the way a particular board expects.
Getting Started: What to ask
If you’re curious about where AI agents could make the most difference, start by asking:
- Where are our people doing the same manual steps, week after week?
- Where do errors slip in that cost time, money, or reputation?
- Which reports or insights do leaders need faster than we can currently deliver?
The answers to these questions often highlight valuable opportunities.
Closing comments
Although we’re hearing the term “AI efficiency” in relation to large corporate layoffs (mostly overseas), for many New Zealand companies the focus is not on replacing their team, but on making their team more valuable.
This approach enables a company to grow its revenue, without having the same commensurate increase in costs.
Ultimately, whatever we think of the use of AI to boost productivity and efficiency, the simple truth is that the competitive advantage it delivers means most businesses will need to adopt these technologies to remain competitive.
But right now there’s an opportunity to be less reactive than that; companies that move ahead during this early phase of AI adoption will be able to accelerate ahead of competitors that lag. This is not only important at a company level, but for national competitiveness on the world stage.
The team at Putti would love to hear your ideas about what boring things in your company could be revolutionized using AI agents.