Tech-First SaaS Investment Tools Platform

Overview

A ground-up re-architecture of a financial tools platform, moving from a bloated monolithic marketplace model to a lean, computation-first SaaS that serves algorithmic calculations and real-time financial data to professional investors. The directive was simple: maximize throughput and precision; minimize operational complexity.

Architecture Highlights

Calculation Engine

At the core is a Python-based computation layer that evaluates multi-variable financial models (DCF, comparables, momentum overlays) on demand. Query results are cached at the PostgreSQL query level with materialized views refreshed on configurable schedules, keeping p99 response times under 80 ms for complex calculations.

API Design

A hypermedia-informed REST API exposes all tools to both the web frontend and potential third-party integrations. Endpoints are versioned, rate-limited per subscription tier, and secured via JWT with refresh-token rotation.

Data Visualization Pipeline

Server-side aggregation pre-computes chart datasets so the client receives ready-to-render JSON — eliminating heavy computation inside the browser and keeping the UI responsive even on low-powered devices.

Infrastructure

Fully containerised with Docker Compose. Nginx reverse-proxy with upstream health checks, automatic TLS renewal via Certbot, and structured logging piped to a lightweight log aggregator.

Stack

Python · Flask · PostgreSQL · SQLAlchemy · Docker Compose · Nginx · Certbot · RESTful API · JWT Auth

Outcome

Platform API response times reduced by ~65% versus the prior monolith. Subscription conversion improved significantly as users could access tools without friction-heavy onboarding flows. The architecture now supports horizontal scaling of the computation workers independently of the web layer.