A Next.js SaaS platform built for a fund manager — live portfolio data feeds, AI-generated performance summaries, and investor reporting automation. Monthly reporting time cut from 8 hours to 20 minutes. Built and delivered to production within 6 weeks.
A fund manager overseeing multiple portfolios was spending 8+ hours every month manually compiling performance data from disparate sources into investor reports. Positions were tracked in spreadsheets, data feeds were pulled manually from broker APIs, and summary narratives were written from scratch each reporting cycle. The process was slow, error-prone, and not scalable as the number of investor accounts grew.
Unified live portfolio and P&L data from multiple broker APIs and data feeds into a single view
AI-generated executive summary that reads naturally with bespoke fund tone, avoiding generic templates
Investor-ready PDF generation rendered on-demand complete with charts, metrics, and narrative commentary
Multi-user access separating full manager admin privileges from secure, read-only investor portals
Fully hosted and automated infrastructure that runs reliably without daily developer involvement
PSI Square built a multi-tenant Next.js SaaS platform with three integrated layers: live data ingestion, AI narrative generation, and investor-ready report export.
Server-side data pipeline pulling live position and P&L data from broker APIs into Supabase on a configurable refresh interval. Normalised across multiple account types and base currencies into a single unified portfolio view.
OpenAI API integration: each reporting cycle, the platform compiles portfolio metrics into a structured prompt and generates a natural-language performance summary tailored to the fund's reporting tone. Manager reviews and edits the AI draft in an inline editor before publishing — AI handles the first draft, human approves the final.
Role-based access: fund manager sees full admin view; investors see only their own portfolio slice and published reports. Features one-click PDF export of the full report including interactive charts, AI narrative, and period performance tables.
System architecture specifications encompassing Next.js server components, OpenAI model routing, and workspace isolation.
Measurable productivity improvements and operational efficiency verified across multiple fund reporting cycles.
Monthly reporting time dropped from over 8 hours to under 20 minutes per cycle. The AI draft handles the narrative, the automated pipeline handles the data, and the manager's time is now spent reviewing and approving rather than compiling. The platform has since been expanded to cover additional fund accounts without any structural changes to the codebase.
We design and engineer enterprise-grade SaaS platforms, AI workflow automations, and custom dashboards.
Whether you need an AI-powered SaaS or a custom internal tool — tell us what you're building. We'll scope it, design it and deliver it to production standard. Most projects start within 5 business days.
Or email info@psi-square.net · Phone +44 (0) 20 3872 7195 · We reply within 24 hours · NDA available on request
