AI Knowledge Search for an Education Group
AI-powered search across policies, procedures, training materials and compliance documents. Staff find accurate answers in seconds instead of searching shared drives and intranets.
Education group — multiple RTOs and training businesses
An education group operating 4 RTOs and 2 training businesses across WA. Over 200 staff, 5,000+ policy and procedure documents, training materials, compliance records and operational guides spread across SharePoint, shared drives and legacy intranets.
Staff spent significant time searching for the right document. Compliance auditors needed specific policies. Trainers needed unit-specific resources. Admin staff needed process guides. Finding the right document — or confirming it was the current version — could take 15–30 minutes.
What needed to change
Documents were scattered across systems. Each RTO had its own SharePoint, shared drives and legacy folders. Some documents were duplicated across locations with different versions. Nobody was confident they were looking at the current version of a policy.
Staff could not find what they already had. A compliance assessor would ask "what is our policy on RPL?" and get 12 results from keyword search — most outdated or from the wrong RTO. They would spend 20 minutes determining which was the current, correct document.
New staff and audits were particularly painful. Onboarding a new trainer meant pointing them at a shared drive and hoping they found the right resources. When auditors asked for specific policies, the compliance team scrambled to locate and verify them.
What we built
An AI knowledge search system using RAG architecture. Staff ask questions in natural language and get synthesised answers from the organisation's policy, procedure and training document library.
Document Ingestion
Automated pipeline indexing documents from SharePoint, shared drives and legacy systems. Version detection ensures only current documents are served. Processes PDF, Word, PowerPoint and Excel.
Natural Language Search
Staff ask questions like "What is our RPL policy for Certificate IV qualifications?" and get a clear answer synthesised from the relevant documents — not a list of files.
Source Attribution
Every answer cites the source document, section and version. Staff can click through to the original document for the full context.
RTO-Aware Results
Search is context-aware — a trainer at RTO A gets results from RTO A's documents first, with group-level policies clearly indicated. Cross-RTO search available when needed.
How it works
Staff types a question
Natural language query — "What are the assessment requirements for CHC33015?" or "Where is the student grievance procedure?" or "What are our trainer qualification requirements?"
System retrieves relevant documents
Semantic search finds the most relevant document sections across the entire library. RTO-specific results prioritised based on the user's organisational context.
AI generates a clear answer
GPT-4 synthesises an answer from the retrieved documents. Constrained to source material to prevent hallucination. Answer includes key points and relevant details.
Source citations provided
Each part of the answer links to the specific source document and section. Document version and last-updated date displayed for confidence.
User drills into full documents
One click to open the source document. Related documents and policies surfaced alongside the answer for broader context.
Measurable outcomes
During our last audit, the assessor asked for 15 different policies. Instead of the usual scramble through shared drives, our compliance manager typed each question into the system and had the answer with the correct document in under 10 seconds. The auditor was impressed.
How we delivered it
Document Audit
1 weekCatalogued all document sources across 4 RTOs and 2 training businesses. Identified the most-searched document categories (compliance policies, training guides, forms). Cleaned up version conflicts before ingestion.
Ingestion Pipeline
2 weeksBuilt the document processing pipeline with multi-format extraction, chunking optimised for policy documents, and embedding. Configured RTO-aware metadata tagging.
RAG Engine & UI
2 weeksBuilt the retrieval and generation pipeline with source citation. Developed the search interface with RTO context switching, document preview and related document suggestions.
Testing & Launch
1 weekTested with compliance, training and admin teams across all RTOs. Refined retrieval quality and answer formatting based on real queries. Launched group-wide.
Similar Project?
Want something similar for your business?
Tell us about your industry, your workflows, and what you want to achieve. We will scope it, quote it fixed-price, and build it.
Tell Us About Your Project
What industry are you in? What systems do you use? What is the biggest operational problem you want solved? We will come back with a plan and fixed-price quote.
Prefer a quick chat? Call 0425 531 127 – we're Perth-based and we answer the phone.