The Challenge
A consulting workflow built around weeks of manual synthesis.
A global advisory firm whose clients include Big Four and top-tier consulting firms had a nine-person consulting team spending weeks manually reviewing 40 to 50 stakeholder interview transcripts per project.
Each engagement required pillar-based intelligence reports that were thorough but painfully slow to produce. The bottleneck was not a lack of talent. It was the sheer volume of unstructured information that needed to be synthesised into structured, actionable insight.
The team's best people were doing work that was necessary but non-differentiating: reading, organising, cross-referencing, and drafting from scratch before any real consulting judgement could begin.
What We Built
An AI report generation engine grounded in the firm's own methodology.
60x built a custom AI report generation engine backed by an enterprise knowledge graph of approximately 10,000 ingested files across 70 projects, with 13 bespoke metadata tags designed around the firm's proprietary analytical framework.
The system ingests interview transcripts, research documents, and internal notes, then generates first-draft observations, implications, and recommendations while preserving source traceability and the client's own analytical structure.
Knowledge graph foundation
Approximately 10,000 files indexed across 70 projects in a single retrieval layer.
Firm-specific logic
13 custom metadata tags mirrored the client's own analytical framework rather than a generic template.
Traceable outputs
Every generated insight can be verified against the underlying source material.
Instead of asking consultants to trust generic AI drafting, the system was designed to reflect how the firm's senior people already structured problems, surfaced themes, and moved from raw evidence to client-ready recommendations.
The Results
From blank-page drafting to substantive first drafts in minutes.
Report drafting time
Projected concurrent projects per year
Sources consulted per query
- Report drafting time: from one full week of manual work to structured first drafts generated in minutes.
- Projected capacity: from 10 to 15 concurrent projects per year without adding headcount.
- Source confidence: every AI-generated recommendation links back to original evidence.
The change was not just speed. It shifted the consultant's role from information compiler to critical reviewer, allowing more time for judgement, challenge, and client-specific nuance.
How It Landed
The system felt useful because it improved the firm's existing method rather than replacing it.
Adoption came from fidelity. The tool did not impose a new workflow or ask the team to think in a generic prompt language. It encoded the firm's actual framework, terminology, and evidence expectations, which made the output immediately legible to consultants.
That meant the first drafts were not simply faster to produce. They were easier to critique, easier to trust, and easier to carry through to final client delivery.
The Compounding Effect
The system improves as the firm does more work.
The knowledge graph does not stop being useful after the first report. It continuously ingests new documents as they land, so every new project, interview, and internal brainstorm makes the system richer and more context aware.
Leadership now views the platform as essential infrastructure rather than a point solution. Plans are already underway to extend the same foundation across executive search, business development, and internal operations.
In practice, this turned one delivery acceleration project into a reusable intelligence layer for the wider firm.