Project Overview
SASLAB designed and built a commercial real estate investment analysis platform that turns complex, multi-tenant cash-flow projections and portfolio management into a fast, reliable web application. A deterministic financial engine sits at the core as the single source of truth, so every figure shown in the interface and in exports stays consistent and auditable.
What We Built
- Investment Analysis: Multi-tenant cash-flow projections, rent rolls, and portfolio-level roll-ups.
- Portfolio Management: Manage buildings, tenants, and leases across an investment portfolio.
- Document Processing: Lease and PDF document handling with report and export generation.
- Payments & Accounts: Subscription and payment handling with secure account management.
- Multiple Interfaces: A customer web application, an administration console, and a public marketing site.
AI Capabilities
The platform's AI features speed up analysis without ever putting the numbers at risk. The rule is strict: the AI reads, interprets, and explains — it never computes a financial figure itself. It parses documents and intent, calls the deterministic engine, and explains the result, so every number stays consistent and auditable.
- Document Abstraction: Reads lease and financial PDFs, including scanned documents through a vision/OCR path, and extracts key commercial terms to populate rent rolls, buildings, and portfolios.
- What-If Advisory: Ask in plain language how a change would affect the numbers, and see the real impact computed by the engine.
- Source-Aware & Grounded: Answers reference the documents behind them, and every figure comes from the financial engine rather than the model.
- Gap Detection: When source documents or amendments are missing or out of date, the assistant reports the gap instead of guessing.
Technology
- Backend: Java 17 and Spring Boot 3, Spring Data JPA / Hibernate, and MySQL, with a REST API documented using OpenAPI.
- Security: Spring Security with JWT-based authentication and role-based access.
- Integrations: Stripe for payments, SMTP email, and Google Maps for property locations, with PDF generation and parsing.
- AI: Large-language-model document abstraction and advisory, orchestrated so the model calls the deterministic engine and never generates financial figures itself.
- Reliability: Error monitoring with Sentry and cloud deployment on Google Cloud (GCP).
- Frontend: React with TypeScript, built with Vite and a component design system.
Engineering Highlights
A deterministic cash-flow engine is the authoritative source for every calculation. The user interface, document exports, analysis tools, and AI features all read from the same engine, which keeps results consistent, testable, and traceable across the platform.
