Skip to content
Client Target IntIndustry SaaS / MicroservicesEngagement Embedded, 18 mo +Timeline Jun 2021 - Jan 2023Status Delivered

A Python microservices SaaS, architected and built.

Over an 18-month engagement we worked as lead and principal engineer on a production Python microservices SaaS platform, owning the architecture as much as the code: a FastAPI and WebSockets service layer, Redis pub/sub messaging between services, and a Docker-based deployment that a growing team could operate and extend.

1,614 hrsDelivered as lead engineer
18 mo +Embedded engagement
5.0Upwork client rating

The challenge.

Target Int was building a production SaaS platform and needed more than a pair of hands. They needed someone to own the technical direction: how the services talked to each other, how real-time updates reached the client, and how the whole thing would scale as the product grew. The work spanned the full stack but the centre of gravity was the backend architecture, the part that decides whether a SaaS stays fast and maintainable or slowly turns into something nobody wants to touch.

Our approach.

We took on a lead and principal engineer role, shaping the architecture and building the core of it.

  • Microservices, not a monolith. The platform was split into focused Python services so each piece could be developed, deployed, and reasoned about on its own.
  • Real-time by design. A FastAPI and WebSockets layer pushed live updates to the client instead of relying on polling.
  • Messaging between services. Redis pub/sub carried events between services, keeping them decoupled while still coordinated.
  • Reproducible environments. Docker-compose described the whole stack so it ran the same way locally and in production.
System Topology, simplified
Real-time APIFastAPI - WebSockets
MessagingRedis pub/sub between services
ServicesFocused Python microservices
RuntimeDocker-compose - reproducible environments

Technology stack.

A conventional, operable Python stack that a team could hire for and extend.

Services: FastAPI, WebSockets, Python microservices.

Messaging and runtime: Redis pub/sub, Docker-compose.

Outcomes.

  • Sustained technical leadership. 1,614 hours over 18 months as lead and principal engineer, not a short fix.
  • An architecture built to scale. Decoupled services with real-time messaging, designed to grow with the product.
  • A platform a team could own. Conventional tooling and reproducible environments, easy to operate and extend.
Shubham is a fantastic backend developer. Expert in Python and Django particularly. Communication was great.
Rohit Thakral · Target Int

Need an engineer to own the architecture?

If you are building a SaaS platform and need someone to shape the backend, not just write tickets, we can lead the architecture and build the core of it with you.