Makena Capital Cuts Analyst Reporting Time by 50% and Builds a Firm-Wide Data Culture on Sigma

Makena Capital's analysts were spending up to half their working hours building static reports from disconnected data sources. Sigma replaced that manual, error-prone workflow with a live, governed analytics layer on Databricks, freeing analysts to focus on investment decisions instead of data assembly.

• 50% – Weekly analyst hours reclaimed from manual report building
• 5 – Net-new workbook capabilities for portfolio projections, client-specific reporting, and interactive exploration
• $20B – Assets Under Management (AUM) supported by Sigma-governed data with penny-level accuracy requirements

“We want analysts to do analysis. We don't want them spending their time building reports. We want them using those reports to make their job better, to deliver more sophisticated insights."
_{— Kunal Koppula, Head of Data Engineering}

The Challenge

Makena Capital is a private endowment managing approximately $20 billion in assets across a fund-of-funds structure. Analysts were spending 10 to 20 hours per week downloading CSVs from disparate data vendors, manually assembling Excel files, and dragging formulas across spreadsheets to produce quarterly reports. The firm's core private assets dashboard was a single, monolithic Excel file with a 50% failure rate on open. No one above the analyst level was willing to operate inside it.

The cost of this model was not just time. Four analysts performing the same data join independently produced four different outputs. In a client-facing business where a 1% return discrepancy triggers immediate scrutiny, inconsistent joins created direct reputational and fiduciary risk. With 20 years of historical data across multiple vendors and no single source of truth, the firm had no reliable way to validate that the numbers in a client presentation matched the numbers in the underlying model.

The Solution

Makena Capital selected Sigma for its native integration with Databricks and its ability to bring non-technical users directly into the analytics layer without requiring engineering support. The firm's data engineering team built governed data models as a shared foundation. Analysts then built workbooks on top of those models without needing to re-derive joins or validate upstream logic themselves.

Sigma's Input Tables and controls became the engine for Scenario Lab, a forward-looking portfolio modeling tool. Analysts can now adjust assumptions for market returns, asset class allocations, and cash flow projections in real time. Conditional formatting and lineage tracking replaced manual audit processes, flagging data anomalies at the transaction level and showing exactly how each number was derived. The firm also uses Sigma as a data validation layer, monitoring ingestion thresholds across raw and cleaned Databricks tables before any number reaches a report.

The Results

Analysts who previously spent 10 to 20 hours per week on report assembly now spend that time on analysis that drives action and client engagement. The private assets dashboard migration empowered investment analysts to become builders themselves, creating self-service tools for portfolio projections, client-specific reporting, and interactive exploration without requiring data engineering support that would not have been possible in the previous Excel paradigm. Penny-level data fidelity is now enforced systematically, not manually, across a $20 billion Assets Under Management (AUM) book where client-facing discrepancies carry direct business consequences.

“We went through an exercise of figuring out how much time our analysts spend building reports. It can be 10 hours a week. It can be 20 hours a week. It's a significant portion of your weekly hours."
_{— Kunal Koppula, Head of Data Engineering}

Makena Capital is now building a self-service client dashboard that consolidates the 15 to 20 most common client questions into a single, permissioned workbook. The firm is also piloting Sigma's Python integration within Scenario Lab to run quantitative models directly on live portfolio data. As AI-assisted workbook building matures, the team expects to onboard new analysts into Sigma within days, compressing the time from question to decision across every investment and client-facing function.