PwC accelerates enterprise-scale GenAI adoption with CrewAI, boosting code-generation accuracy from roughly 10% to 70%+
PwC consultants needed faster, more-accurate generation of proprietary-language code and lengthy spec documents, but early Gen-AI prototypes produced inconsistent results and offered little transparency into ROI, undermining user trust.
How it works
Common implementation structure
How this type of workflow is generally built, generalized across documented cases — not tied to any one vendor's stack. Click any stage to read what happens there. Specific products that implement these stages appear in “Tools commonly seen” below.
Stage 1 · Consultant requests code or document
PwC consultants initiate requests for proprietary-language code and lengthy spec documents.
Tools used
Crew AI
Outcome
Crew-powered agents boosted code-generation accuracy from 10% to 70%+, slashed turnaround time on complex documents, and supplied granular ROI data, restoring consultant trust and accelerating adoption of agentic solutions across PwC.
What failed first
PwC initially built its own plug-in framework during its firm-wide Gen-AI transformation, but the early prototypes lacked real-time feedback, produced inconsistent results at around 10% accuracy, and offered no transparency into ROI.
Results
Time savedslashed turnaround time
Volumeroughly 10%
Running sincetwo years ago (relative to publication)