The Challenge
Strata Finance is a mid-market asset management firm with £1.2bn AUM. Their monthly board pack — a 40-slide PowerPoint deck covering portfolio performance, risk metrics, fee income, and client activity — was built manually by two analysts. The process took two full working days each month.
Worse, by the time the pack reached the board it was already five days old. Decision-making was based on data that was effectively a fortnight behind reality.
The data problem
Their KPIs lived in six different systems: Xero (financials), Salesforce (client relationships), a proprietary risk model in R, a fund administrator portal (CSV exports only), Bloomberg (market data), and a compliance platform. Getting them into one place required manual extraction, formatting, and reconciliation every month.
Our Solution
We built a centralised analytics platform with automated data pipelines feeding a real-time dashboard, plus a scheduled report engine that generates the board pack as a branded PDF on Monday mornings without human intervention.
Six-Source Data Integration
Live API connections to Xero and Salesforce. Scheduled imports from the fund administrator via SFTP. Bloomberg market data via their API. R model outputs piped through a REST wrapper we built. All data normalised into a unified schema.
Real-Time Executive Dashboard
Role-separated views: portfolio managers see performance and attribution; operations see fee income and client activity; the board sees the top-line KPIs. Charts built with D3.js, refreshed every 15 minutes during market hours.
Automated Board Pack Generation
Every Monday at 06:00 GMT, the system generates a branded PDF board pack pulling the latest data, applies Strata's house style, and emails it to board members and their assistants. Analysts review rather than build.
Threshold Alerts
Configurable alerts fire via email and Slack when key metrics cross defined thresholds — drawdown limits, client concentration breaches, fee shortfalls. No need to monitor dashboards manually.
Technical Approach
The biggest challenge was the fund administrator's data — only available as CSV exports with inconsistent column naming across months. We built a normalisation layer that uses pattern matching and fuzzy column matching to handle this reliably, with manual override and audit logging when automatic matching fails.
The PDF generation engine uses Puppeteer to render a React template as HTML, then captures it as a high-fidelity PDF. This gave us full control over layout and branding without the limitations of a dedicated reporting library.
Technology Used
Results
"We replaced three separate SaaS subscriptions with one custom platform. ROI in under 8 months. The team at Webmatx genuinely understood finance, which made all the difference."
