AI Platform Consultant (Career Break)
Independent
Nov 2022 - Jun 2023 (8 months)
• Authored a private Turbo‑ AI template that bundles FastAPI, SQLModel, LangChain, pgvector, Redis, Docker Compose, and Kubernetes manifests. • GitHub Actions logs show a median time of forty‑ six minutes from cloning the template to a functional chat and retrieval augmented generation demo. • Implemented pluggable vector adapters for Pinecone, Weaviate, and Milvus so that teams can migrate seamlessly when scale demands. • Replaced Celery version five job queues with Temporal workflows for transcript parsing and nightly retraining, and no ghost task incidents were observed during a six‑ week soak test. • Provisioned preview stacks for every pull request with Terraform modules and GitHub Actions, and GitHub Insights reports a ninety‑ fi