Traditional reporting tools provide valuable benchmarking analysis but are primarily focused on hindsight. Carve takes reporting a step further by providing foresight. Instead of only reviewing results after the fact, Carve integrates AI summarization and benchmarking insights directly into the gate review process. This enables project teams and reviewers to communicate insights and results in a more meaningful way. Power BI is effective in reporting history, while Carve Hybrid Intelligence is designed to inform decisions before they are made.
Great facilitators share a few key traits.
Carve supports facilitators by streamlining the process, reducing administrative effort, and enabling more consistent, successful outcomes.
The duration varies by template and project size. For PDRI Industrial, one would typically need a full-day session for mid-sized projects, and longer for mega projects. For PDRI Small Industrial or Small Infrastructure, one would need around 2-2.5 hours. For PDRI Buildings, one would need about 4 hours.
Each template differs in complexity and depth, but all are designed to deliver consistent, reliable outcomes.
Company benchmarks are created from your own completed and published PDRI Assessments. Administrators have the option to exclude outliers (e.g., joint ventures or atypical projects) to ensure benchmarks accurately reflect your organization’s processes. These benchmarks can then be analyzed across your portfolios.
Industry benchmarks are based on aggregated data from Carve customers who opt in. To be included, there must be at least 10 assessments at the same stage across three different companies. Carve applies rigorous data quality checks, such as filtering invalid or duplicate score and removing statistical outliers, to create a clean, high-integrity dataset. Today, the largest industry dataset is for PDRI Industrial templates, which now include thousands of assessments.
Carve supports portfolio assurance by providing a centralized system where all projects, regardless of size or complexity, can be evaluated using Construction Industry Institute (CII) templates. This allows assurance teams and gate reviewers to access standardized data across the portfolio, ensuring a consistent, one system approach. It also streamlines onboarding and training, as everyone uses the same methodology and platform.
Carve separates benchmarking from AI summarization to ensure accuracy. Benchmarking is handled through structured comparisons of scores against company or industry datasets; AI is not used for this, eliminating the risk of ambiguity or “hallucinations.” AI is used only to summarize comments and action items, after which benchmark comparisons are appended as factual statements.
Yes. Benchmarking results are contextualized by stage-gate. For example, scores should be higher at Gate 2 and progressively lower by Gate 3 as project definition improves. This gate-based progression provides meaningful insight into how well projects are advancing.
No. Carve does not send any customer data to ChatGPT or any external AI system. Instead, Carve uses a large language model (Llama, hosted within Amazon Web Services) that runs entirely inside its secure cloud environment. Data never leaves the Carve application, nor is it ever used to train any model. Each AI-generated summary is created for the specific assessment at hand, ensuring the highest levels of confidentiality and compliance with data residency requirements.
Carve Hybrid Intelligence was released July 4, 2025. Early assessments are only now beginning to pass through gate reviews, so quantitative data is still being collected. However, early qualitative feedback from customers has been very positive, with more concrete results expected as adoption continues.
The Hybrid Intelligence features for construction readiness, such as AI-generated summaries and insights, are in active development and scheduled for release by the end of Q3 2025. These insights will differ slightly due to the unique structure of the construction readiness assessments but will include the same consistent executive summaries and aligned reporting.