Backend & API Engineering
Versioned REST APIs with Django REST Framework and FastAPI, documented with OpenAPI, validated with Pydantic and kept thin through a clean service layer.
Tech Lead at TerrificMinds · Open to senior backend conversations
SreeshanthShimji
I'm aSenior Python Developer
I design and build secure, scalable Python backends — REST APIs, multi-tenant SaaS platforms and cloud infrastructure — with Django, FastAPI and AWS. 8+ years, from Kochi to clients in Australia, the USA and the Philippines.
~1 Lakhtenants served
8+ yrsPython backends
PythonDjango 5Django REST FrameworkFastAPIPostgreSQLMySQLRedisCeleryRabbitMQDjango ChannelsAWS ECS FargateAPI GatewayDockerTerraformGitHub ActionsDatadogpytestFHIRPythonDjango 5Django REST FrameworkFastAPIPostgreSQLMySQLRedisCeleryRabbitMQDjango ChannelsAWS ECS FargateAPI GatewayDockerTerraformGitHub ActionsDatadogpytestFHIR
Impact in numbers
Backends that carry real load.
0+
Years building production backends in Python
0 Lakh
Tenants served by one multi-tenant API platform (approx.)
0
API endpoints designed & shipped on one platform (approx.)
0+
Organisations on a learning platform I led
What I do
End-to-end backend engineering, from schema to production.
I take ownership of the parts of a product that have to be right: the data model, the API contract, security, and the pipeline that ships it.
Multi-tenant SaaS Architecture
Tenant isolation, organisation-based request routing and per-tenant database resolution for platforms with tens of thousands of customers.
Data Modelling & Databases
Schema design, migrations (Django, Alembic) and query optimisation across PostgreSQL, MySQL, Sybase SQL Anywhere and Redis.
Cloud & DevOps
Containerised services on AWS ECS Fargate behind API Gateway, provisioned with Terraform and shipped through GitHub Actions CI/CD.
Security & Compliance
JWT and OAuth2, API keys, RBAC, MFA, rate limiting and audit logging — plus hands-on work on ISO 27001 engineering practices.
Real-time & Async Systems
WebSockets with Django Channels, Celery and RabbitMQ pipelines, webhooks and scheduled jobs that keep data flowing reliably.
How I build
Thin views. Explicit services. Tests that gate the merge.
Every request is authenticated, authorised, rate-limited and routed to the right tenant before a single query runs — and every change is reviewed and covered by CI before it reaches production.
Capability-based accessPer-tenant DB routingRate limiting80% coverage gateDatadog tracing
# Resolve every request to the right tenant database @router.get("/v1/{org_id}/appointments") async def list_appointments( org_id: str, key: ApiKey = Depends(require("appointment:read")), ): await rate_limiter.check(key, per_minute=60) db = tenant_router.resolve(org_id) return await AppointmentService(db).list() # $ pytest --cov --cov-fail-under=80 ========== 412 passed · coverage 86% ==========
Illustrative snippet showing the pattern — not production code.
Selected work
All projects Featured case studies
Healthcare integration · Multi-tenant SaaS
Centaur Dental4Web API Platform
A multi-tenant integration platform that lets third-party apps securely read and write dental practice data — ~1 lakh tenants, thousands of integrators, ~100 endpoints.
PythonDjango 5Django REST FrameworkFastAPISQLAlchemyAlembic
~1 Lakhtenants (practices) served
1000sthird-party integrators
~100API endpoints
01
Integrator requestPartner app calls the versioned public API with an organisation ID
02
AWS API GatewayAPI key authentication and usage plans at the edge
03
Capability & rate-limit checkTier capabilities plus per-minute and daily limits
04
Organisation routerResolves the org ID to the correct per-practice Sybase database
05
FastAPI service layerPydantic validation, thin services, FHIR mapping
06
Versioned responseOnly the data that key is allowed to access
EdTech · AI roleplay · Real-time
Edustack – First Principles
A multi-tenant training platform for dental teams with real-time AI roleplay over WebSockets, Celery scoring pipelines and privacy-first call ingestion.
Django 5Django REST FrameworkSimpleJWTMySQLRedisDjango Channels
100+organisations
1000slearners
4activity types
training_room.queuemax_unfinished = 8
empty
empty
empty
empty
empty
empty
empty
empty
■ assigned■ live call■ completed → refill
E-commerce · Australia
DSHOP – E-Commerce Admin Platform
The admin dashboard behind a major Australian e-commerce platform — millions of customers, 1,000+ admin staff, orders to RMA to stock.
PythonDjangoPostgreSQLRabbitMQCeleryCron
Millionsof end customers
1,000+admin staff on the dashboard
9operational modules
Let's talk
Need a backend that scales quietly?
I'm open to conversations about senior and lead backend engineering roles, and about hard problems in Python, APIs and multi-tenant SaaS.