Start-up Architecture Assessment
Healthcare
Context
A healthcare technology startup had been unable to deliver new product features for several months. The technical team reported they were fully occupied with production incidents and system stability issues. Board members and investors were pressuring the CTO to keep delivering promised features.
Engagement
We conducted a technical investigation over one week, examining the production architecture, codebase activity, and incident patterns. This included reviewing tickets and source control to understand how engineering time was actually being spent, analyzing the relationships between microservices, mapping data flows across multiple database systems (MongoDB, MySQL), and interviewing team members about operational patterns.
Findings
Our investigation documented a microservices architecture that had become operationally unsustainable. A central routing service handled all inter-service communication, creating both a single point of failure and a performance bottleneck that regularly caused system-wide outages. Analysis of git activity showed engineering time split approximately 75% incident response and 25% feature development. We identified systematic data integrity issues resulting from the use of MongoDB for transactional data without proper consistency controls, leading to regular synchronization failures with MySQL systems and documented data loss incidents affecting customer records. The architecture reflected design decisions inconsistent with distributed systems best practices and standard approaches to referential integrity.
Outcome
The board received a factual assessment of technical viability and specific areas requiring intervention. They were able to make informed decisions about leadership and architectural direction. They onboarded a fractional CTO to assist the titular CTO, which resulted in restored feature delivery capability and reduced operational burden.