Montagu: Reporting cycle from three weeks to two days

The problem

Fifteen portfolio companies. Each reporting monthly financials in different formats, different systems, different cadences. The group finance team spent three weeks per quarter manually aggregating it all into an investment committee pack. By the time it landed, the data was already stale.

The process required too many manual handoffs between systems that did not speak to each other. Every quarter, someone discovered a formula error in a submission, and the correction cascaded through every downstream calculation.

The solution

We built a data pipeline that ingests portfolio company data regardless of format, normalises it into a common schema, and feeds it into an analytics layer. Adaptive parsers learn each company's reporting format rather than forcing standardisation at the source. Anomaly detection catches issues before they reach the investment team.

The system runs inside Montagu's environment, on their infrastructure, with their access controls. No data leaves their network.

The result

The reporting cycle dropped from three weeks to two days. Day one is automated ingestion and analytics. Day two is human review and sign-off.

The investment committee now sees portfolio performance in near real-time rather than through a quarterly rearview mirror. Analysts who previously spent the majority of each quarter aggregating data now spend it on analysis.