AI Knowledge Search for a Consulting Firm
AI-powered search across 10 years of reports, proposals and internal knowledge. Consultants find relevant past work in seconds instead of searching shared drives.
Management consulting firm — strategy, operations and advisory
A 50-person management consulting firm with 10 years of accumulated knowledge — client reports, proposals, industry research, methodology documents, case studies and internal processes. Over 15,000 documents across SharePoint, shared drives and individual folders.
Consultants frequently duplicated work because they did not know similar work had been done before. A senior associate might spend 2 days researching a topic that another team had covered comprehensively 18 months ago — the report existed, but nobody could find it.
What needed to change
Institutional knowledge was invisible. 15,000+ documents accumulated over a decade, but no way to search them effectively. SharePoint search was keyword-based and returned irrelevant results. Most consultants relied on asking colleagues rather than searching.
Work was duplicated constantly. The firm estimated that 20% of research and analysis work was duplicated across projects — a methodology framework developed for one client would be rebuilt from scratch by another team that did not know it existed.
Knowledge walked out the door. When senior consultants left the firm, their knowledge — which projects they had worked on, what approaches they had used, what they had learned — left with them. No way to capture or surface it for the remaining team.
What we built
An AI knowledge search system using RAG architecture. Consultants ask questions in natural language and get answers synthesised from the firm's entire document library, with source citations.
Document Ingestion
Automated pipeline processing documents from SharePoint and shared drives. PDF, Word, PowerPoint and Excel files are extracted, chunked and embedded for semantic search.
Semantic Search
Natural language queries — "What frameworks have we used for supply chain optimisation?" — return synthesised answers from relevant documents across the entire library.
Source Citations
Every answer links to the source documents and specific sections. Consultants can verify answers and access the full original documents.
Knowledge Graph
Automatic tagging of topics, clients, industries and methodologies across the document library. Related documents surfaced alongside search results.
How it works
Consultant types a question
Natural language query — "What did we recommend for the logistics client on fleet optimisation?" or "Do we have any carbon reporting frameworks?"
System retrieves relevant knowledge
Semantic search finds the most relevant document chunks across the entire library — reports, proposals, methodologies, meeting notes.
AI synthesises an answer
GPT-4 generates a clear answer based on the retrieved documents. Constrained to source material to prevent hallucination.
Citations link to source documents
Each claim in the answer links to the specific source document and section. The consultant can click through to read the original.
Related documents surfaced
Alongside the answer, the system shows related documents — other reports on the same topic, relevant proposals, applicable methodologies.
Measurable outcomes
A consultant asked the system what we had done in aged care strategy. It found 3 reports from different years, a proposal template and a methodology framework — all in 5 seconds. That would have taken a day of asking around and searching shared drives.
How we delivered it
Knowledge Audit
2 weeksCatalogued all document sources — SharePoint, shared drives, email archives. Assessed document quality, format variety and access controls. Identified the most valuable knowledge categories.
Ingestion Pipeline
3 weeksBuilt the document processing pipeline. Handled PDF extraction, PowerPoint conversion, table processing and OCR for scanned documents. Optimised chunking strategy for consulting-style documents.
RAG Engine & UI
2 weeksBuilt the retrieval and generation pipeline. Developed the search interface with citation linking, related documents and topic navigation. Implemented access controls matching SharePoint permissions.
Testing & Launch
1 weekTested with a group of 10 consultants across different practice areas. Refined retrieval quality based on real query patterns. Launched firm-wide with training sessions.
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.