Recruitment Process Automation: What to Automate and How
Recruitment process automation means handing the repetitive steps of hiring to software: posting jobs, collecting applications, first-pass screening, scheduling, candidate updates and reporting. The best results come from automating those steps around the tools you already use and keeping people in charge of every decision that affects a candidate.
By Matthew Clarkson · November 18, 2026

Recruitment process automation means handing the repetitive steps of hiring to software: posting jobs, collecting applications, first-pass screening, scheduling, candidate updates and reporting. The best results come from automating those steps around the tools you already use and keeping people in charge of every decision that affects a candidate. We build this kind of automation for Australian businesses, and the sequence below is the one we would give a friend running a hiring team: map the work, automate the admin, keep judgement human, and check the privacy rules before you go live.
Reduce Repetitive Hiring Admin
Osher builds custom automation around your existing tools, helping teams reduce manual data entry and give people more time for higher-priority work.
TL;DR
Automate the admin around hiring (job intake, resume formatting, scheduling, status updates, reporting). Keep shortlisting decisions, rejections and offers with people.
Automated candidate screening works best as one input to a human decision. Australian public-service guidance advises against hard screening decisions based on a single AI assessment.
Privacy law already covers unsuccessful applicants, and new automated decision-making disclosure duties start on 10 December 2026.
Start with an audit of where recruiter hours go, then connect your existing ATS, inbox, calendar and HR tools rather than replacing them.
A custom build you own avoids per-user fees. Our Business Process Audit is from $7,500, one-off.
What recruitment process automation covers
A hiring process has the same shape in almost every Australian business. The government's hiring guide on business.gov.au lists the steps: prepare a job description and selection criteria, advertise the position, assess applications and shortlist, interview the shortlisted candidates, then choose the best one. Automation takes the typing, copying, chasing and formatting out of each of those steps.
It helps to separate three things that are often sold as one:
An applicant tracking system (ATS) is the record of who applied, what stage they are at and what was decided. It stores; it does not do the work between stages.
Recruitment automation is the set of rules and workflows that move information between stages and people: when a candidate reaches interview, book the slot, send the confirmation, update the record, notify the hiring manager.
AI adds a layer that reads, summarises, extracts or ranks information, such as pulling structured details out of a resume. The decision stays with the recruiter or hiring manager.
HR process automation is the wider field. It covers onboarding, leave, payroll, superannuation and compliance reporting as well as hiring. If your pain is the hiring funnel, start there. If your pain is everything after the contract is signed, the same method applies, but treat it as a separate project.
The same discipline of mapping the process before choosing tools applies in other functions too. This sales process automation guide walks through it for a sales pipeline, and the logic carries over to a hiring pipeline.
Which recruitment tasks are worth automating
The rule we use: automate work that is repetitive, rule-based and high volume, and that a person does the same way every time. Leave work that depends on reading a person.
| Hiring stage | Good to automate | Keep with a person |
|---|---|---|
| Job intake | Turning a hiring manager's brief or call notes into a structured job record and draft advert | Agreeing the role, the selection criteria and the salary band |
| Advertising | Posting to job boards and your careers page from one record | Choosing channels for hard-to-fill roles |
| Applications | Capturing applications, extracting resume details, reformatting resumes into one house style, deduplicating | Reading borderline or unusual applications |
| Screening | Checking applications against the stated criteria, flagging missing information, drafting a summary | The shortlist decision |
| Scheduling | Offering interview times, sending confirmations and reminders, rescheduling | Conducting the interview |
| Communication | Acknowledgements, status updates, reminders for referees | Rejections to interviewed candidates, offers, negotiation |
| Reporting | Time to fill, source of hire, pipeline by stage, recruiter workload | Deciding what to change |
Two patterns show up again and again in agency and in-house teams. The first is intake and formatting: someone retypes a job brief, or reformats every resume into a client template, hour after hour. The second is chasing: following up candidates, referees and hiring managers by hand. Both are good first projects because the rules are clear and the saving is easy to measure. You can see how we approached the first on a job intake automation project for a recruitment agency and the second on a resume formatting agent for a medical recruitment agency. Our recruitment and HR automation page sets out the wider scope.
Automated candidate screening: how it works and where people stay in
Automated screening, including automated resume screening, usually does one or more of these jobs: extracts details from a resume into fields, compares them with the criteria for the role, flags missing information, and ranks or groups applicants for a recruiter to review.
The risk is not the speed. It is a single automated score becoming the reason someone is rejected. The Merit Protection Commissioner's guidance on AI-assisted recruitment tools, which covers resume scanners, AI-reviewed video interviews and psychometric tests, advises minimising hard screening decisions based on one AI assessment, using AI as one part of a holistic assessment, and keeping humans involved at different stages to review what the tool recommends. It was written for the Australian Public Service, but the principle suits any employer.
The scale that careful automation can reach is visible at the large end. Kmart Australia's chief people officer told Human Resources Director that its AI assessment tool cut hiring time from 44 days to 11.8 days, screens about 600,000 applications a year, and that human recruiters hire about 85% of its recommendations. Your volumes will be smaller, but the structure is the one to copy: the tool narrows and organises, people decide.
In practice we design screening this way:
Write the criteria first, in plain terms, from the job description and selection criteria.
Have the system extract and summarise against those criteria, and show its reasoning to the recruiter.
Route every borderline case to a person rather than auto-rejecting it.
Keep the recruiter's final decision and reason in your own records.
That last point matters for fairness. If the data a tool learned from reflects past hiring bias, it can repeat that bias. Two Birds, an employment law practice, recommends regular audits of AI systems to find and correct discriminatory patterns. Build that check into the routine from day one.
Moments that should stay human
Some steps are cheap to automate and expensive to get wrong. We would keep these with a person:
Rejections after an interview. A candidate who gave you an hour deserves a human reply, and they are your future referrals.
Outreach to high-value candidates. Generic messages read as generic.
Salary negotiation and offers. The content is a judgement; automation can draft the paperwork afterwards.
Anything unusual. Gaps in employment history, career changes and overseas qualifications are exactly where rigid rules misread people.
Automation then does the surrounding admin: it assembles the interview pack, drafts the offer letter from approved data and updates the record. Your team spends the saved hours on these conversations.
Privacy and legal rules to settle before you automate
Automation does not change your obligations, but it can make them easier to miss. Three points matter for Australian employers.
Applicant data is covered. The Fair Work Ombudsman's workplace privacy guide states that the Australian Privacy Principles under the Privacy Act 1988 apply to personal information about unsuccessful job candidates, including resumes, contact details, references and academic transcripts. It adds that third parties providing recruitment, HR or payroll services under contract may also need to comply. Know where candidate data is stored, who can see it and how long you keep it.
Automated decision-making disclosure starts on 10 December 2026. From that date, organisations covered by the Privacy Principles must state in their privacy policies the kinds of personal information used by computer programs to make decisions that could reasonably be expected to significantly affect a person's rights or interests, and the kinds of such decisions. The Office of the Australian Information Commissioner sets this out in its issues paper on the transparency obligation. That paper also suggests using short, plain-language messages and checking that people understand them, rather than legalistic wording. If a tool screens or ranks applicants, review your privacy policy and application page before that date.
There is no AI-specific hiring law, but existing law applies. Two Birds notes in its Talent Wars guide that Australia relies on voluntary AI Ethics Principles alongside the Privacy Act, and that employers must tell individuals the purposes for which their information is collected, including use in an AI system. Discrimination law still applies to outcomes, whoever built the tool. If you are unsure how your obligations apply to a particular tool, speak to an employment lawyer before it goes live.
How to automate your recruitment process: step by step
This is the routine, with what each step demands and where it tends to go wrong.

1. Map where the hours go
For two weeks, have recruiters and coordinators log time against each stage in the table above. Most teams are surprised: the heaviest load is rarely screening. It is retyping, reformatting, scheduling and chasing. The common mistake is automating what feels slow rather than what is measurably slow.
This is the first thing our Business Process Audit does. The audit is from $7,500, one-off, and prioritises automation opportunities so you build the highest-value ones first.
2. Define the rules in writing
Automation can only follow rules someone has stated. Write down when a candidate moves stage, who approves what, what a complete application looks like and which messages go out at each point. Gaps here become errors later. People often discover during this step that two recruiters run the process differently.
3. Connect the tools you already have
Your ATS, email, calendar, job boards, HR system and accounting tools already hold the data. Integration moves it between them automatically, so nothing is keyed twice. Replacing your ATS is rarely necessary. We connect existing systems such as Microsoft 365 and HubSpot, and our automation and integration work is from $5,000, one-off. The mistake to avoid is a fragile chain of connections nobody documents. Ask for the workflows to be documented and owned by you.
4. Add AI where it earns its place
Resume extraction, summarising, drafting job adverts from a brief and sorting inbound email are strong uses. Our approach is to use AI agents only where they add real value and to retain human judgement in every critical decision. AI agents and assistants are from $4,000, one-off. See our HR recruitment AI agent and AI application processing project for a talent marketplace for examples of the territory.
5. Pilot on one role type
Run the new workflow on one role family, such as high-volume administrative hires, alongside your existing process. Compare the output, fix the rules and only then widen it. The common failure is switching everything on at once and losing trust in the system when the first edge case breaks it.
6. Measure and review
Track a short list monthly:
Time to fill, from advert to accepted offer
Recruiter hours per hire on admin tasks
Candidate drop-off by stage
Share of automated recommendations that recruiters accept or override
Complaints or queries from candidates
The override rate is the most useful of these. If recruiters override constantly, the criteria are wrong. If they never override, check they are still reading.
Custom build or off-the-shelf recruitment software?
Many businesses begin by looking for a recruitment platform. That works when your process is standard. It is harder when you run an unusual intake process, place candidates for clients with their own templates, or work across several systems that were never designed to talk to each other.
A custom system built around your workflow handles those specifics. With our work you own all code, data and IP outright, and there are no per-user fees or forced subscriptions, so the cost does not grow every time you add a recruiter. A full bespoke business system is from $15,000, one-off. For high-volume, rules-based tasks across legacy software, robotic process automation is $15,000 to $60,000, one-off.

Costs and what a good result looks like
Businesses often ask for a number. The honest answer is that it depends on how many steps you automate and how many systems they touch, which is why we start with an audit. What you can do is set the target before you spend: for example, hours of admin per hire, or days from application to first contact.
For context on what time savings can look like, the Kmart figures above show the top end of what a large employer reported. We would not promise a number to a business your size without seeing your process. One of our property clients identified $250,000 in annual savings through automation, and the method is the same wherever manual admin accumulates: find it, price it, remove it.
Where we can help
We automate business processes for Australian organisations, including recruitment, property management, healthcare, legal and professional services. For hiring teams that usually means a mix of integration, AI agents for the admin-heavy steps, and a custom system where the process does not fit a standard tool. We start with a Business Process Audit, from $7,500 one-off, so the work you fund first is the work that saves the most.

If your recruiters spend their days on data entry and chasing, see our services or talk to our AI consultants about a first project. You can also read our view on AI in recruitment.
FAQ
What is the difference between recruitment automation and an ATS?
An ATS records candidates and where they are in the process. Recruitment automation is the set of workflows that does the work between those records: sending messages, booking interviews, moving data and producing reports. Most businesses need both, connected, so information is entered once.
Can I automate resume screening and stay compliant?
Yes, if people stay involved and you are open about it. Australian guidance advises against hard screening decisions based on one AI assessment, and from 10 December 2026 privacy policies must describe automated decision-making that significantly affects people. Use automation to organise and summarise, and have a recruiter make the decision.
Which recruitment tasks should I automate first?
Start with high-volume, rule-based admin: job intake, resume formatting and extraction, interview scheduling, candidate updates and reporting. They are easy to measure and carry the least risk to candidates. A process audit will show which of these costs your team the most hours.
How long does a recruitment automation project take and what does it cost?
It depends on the number of steps and systems involved. Our published starting prices are $4,000 for AI agents and assistants, $5,000 for automation and integrations, and $7,500 for a Business Process Audit, all one-off. We scope larger builds after the audit.
Sources
Guidance material for using AI-assisted recruitment tools, Merit Protection Commissioner
Kmart's 'blind' AI tool cuts hiring time and costs, Human Resources Director
Automated Decision-Making Transparency Obligation (APP 1), OAIC
Recommended
Last updated on November 18, 2026
Free tool · 2 minutes
Automation ROI Calculator
See what a manual process is really costing you, and the payback on automating it.
Continue Reading
View all
SafetyCulture (Mitti) vs Fulcrum vs Custom Software for Inspection Reports
When the report is what you sell, form builders hit limits. How Mitti (formerly SafetyCulture), Fulcrum and owned software compare on report output, offline capture, hosting and cost.
Oct 27, 2026

n8n vs Power Automate vs Make: Which Fits a Microsoft 365 Business?
A practitioner comparison for Australian businesses on Microsoft 365: what you already pay for, how each tool bills, where the data sits, and what happens when a flow fails.
Oct 27, 2026

Rented software was always a compromise. AI has ended it.
Off-the-shelf software made sense when custom cost $250,000. AI-assisted development changed the maths. Why we now build software around the business, AI-first, and connected to Claude and ChatGPT, starting at $15k.
Aug 6, 2026
Sick of reading about automation?
Book a free 15-minute intro call. We’ll talk through what you’re trying to automate and whether we’re a good fit.