Case Study — HR Learning & Community Platform

100+ tables. One monolithic database. Rebuilt for the next phase — hybrid PostgreSQL + DynamoDB on AWS.

How an HR learning platform with 100+ database tables moved from a monolithic architecture to a hybrid PostgreSQL/DynamoDB system built for the next phase of growth.

100+

Tables assessed in schema analysis

Hybrid

PostgreSQL + DynamoDB architecture

AWS

Native serverless migration strategy

Phased

Implementation with no feature freeze

Before.

The platform had outgrown the database it launched on. 100+ tables in a monolithic structure that was never designed for high-volume event data alongside relational content, user, and community data.

Every new feature added friction to the same structural bottleneck. Aging Node.js 14 infrastructure and lack of real-time capabilities limited the platform's ability to support its growing 73,717 registered users and 12,575 monthly actives.

The Situation

Monolithic PostgreSQL database handling all data types
Technical debt from 100+ tables with unused columns
Scalability limitations during peak event loads (371K sessions)
No dedicated infrastructure for high-volume social activity logs

What we did.

Architected a hybrid data strategy for scale and velocity.

Step 1 — Schema Analysis

Assessed 100+ tables, categorising data by type: relational (users, content, events), high-volume (activity logs, event streams), and transactional.

Step 2 — Architecture Decision Record

Evaluated hybrid approaches; selected PostgreSQL for relational data and DynamoDB for high-volume event data, documenting trade-offs in a formal ADR.

Step 3 — Hybrid Architecture Design

Designed a full AWS-native architecture using ECS/Fargate, RDS, and DynamoDB with API Gateway WebSocket support for real-time features.

Step 4 — Migration Plan

Developed a phased 6-month migration implementation plan using the Strangler Fig pattern to avoid a "big-bang" rewrite.

Step 5 — Developer Implementation Guide

Authored a practical guide for the engineering team, including CI/CD pipelines, infrastructure-as-code templates, and modern ORM adoption.

After.

73,717

Users supported by a scalable hybrid architecture

< 50ms

Target query performance for social features (95th percentile)

6-month

Phased migration timeline with zero feature freeze

73%

Reduction in migration cost vs. original external estimates

99.9%

Target infrastructure stability and platform uptime

+40%

Projected increase in development velocity post-migration

The Installed System.

Phased Migration Roadmap

The engineering team has a migration plan they can execute in phases — feature development continues during the migration using the Strangler Fig pattern.

Hybrid Data Decoupling

Relational data stays relational in PostgreSQL, while high-volume event data moves to DynamoDB. Each database now does exactly what it's designed for.

Scaling Infrastructure

New features no longer require working around structural bottlenecks. The architecture supports the next 3–5 years of growth without another foundational rebuild.

Jake McMahon

ProductQuant

10 years building growth systems for B2B SaaS companies at $1M–$50M ARR. BSc Behavioural Psychology, MSc Data Science. This engagement required performing a deep forensic audit of a 100+ table schema to architect a hybrid SQL/NoSQL system that resolved foundational technical debt while enabling next-gen social features.

Architecture built for launch, not for scale?

A 15-minute call is enough to know whether what we do is relevant to where you are. No pitch. Just a conversation about your specific situation.

Start a Conversation → See All Offers →