Shishir Modi
Building & scaling the infrastructure behind AI systems
Engineering reliability at scale
I specialise in AI platform infrastructure and full-stack development — designing containerised, production-grade systems from inference pipelines to event-driven backends. I bring a disciplined engineering mindset to reliability, observability, and scalable deployment across cloud-native environments.
Infrastructure
Observability
AI / MLOps
Languages
Full-Stack
Selected work
Production-grade systems spanning AI inference, event-driven architectures, and full-stack applications.
Skaitch
End-to-end AI inference platform serving a Stable Diffusion + DeepFaceDrawing pipeline containerised with Docker and orchestrated via Kubernetes with horizontal pod autoscaling (HPA).
Crowd Psychology State Engine
Microservice-based simulation backend using FastAPI and Redis pub/sub, decoupling the probabilistic state engine from downstream consumers via an event-driven message bus supporting concurrent simulation sessions.
RiteFit
Production-grade full-stack application with a React frontend, FastAPI backend, and PostgreSQL persistence layer; integrated live weather APIs to power real-time, AI-driven outfit recommendations with sub-200ms API response times.