- Role: SDE‑3 building and operating production‑grade cloud‑native platforms on OpenShift (OCP) and Rancher.
- GenAI/LLM: Delivered LLM‑powered applications to production (RAG, observability, guardrails, cost/latency tuning).
- Stack: Python (Django/Flask), Docker, Kubernetes, ML (TensorFlow/PyTorch), AWS.
- Focus: Reliability, performance, and developer experience at scale.
- Cloud‑Native
- Cluster and app lifecycle on OCP/Rancher, GitOps, secure CI/CD, and observability.
- Containerized microservices with clear SLOs and production readiness.
- LLM in Production
- Building retrieval‑augmented systems, prompt and embedding pipelines, and monitoring.
- Ensuring safety, evaluation, and rollback strategies for GenAI features.
- cloakprompt-cli: CLI to redact secrets before sending data to AI models (LLM safety, prompt security).
- tf-inference-devops: TensorFlow image classification inference pipeline with Flask + Docker.
- scb-deployments: Helm deployment manifests for cloud‑native workloads.
- Diabetic_retinopathy: Flask + ML app to detect diabetic retinopathy (⭐ starred).
- Real-Fake-Image-Classification: CIFAKE dataset—classify real vs AI‑generated images.
More here: All repositories
