Asdrúbal S
Venezuela
$ 5,700/month
10+ yrs exp
Riley's backend depended on a single LLM endpoint, and uptime sat at 93% because that endpoint was often down or rate-limited. I was assigned to integrate Google Cloud Vertex AI. I went a step past that and built a multi-provider setup, Vertex AI as the primary and Claude as the fallback, and uptime went to 99%. The downtime that was left wasn't coming from the LLM provider, so I increased the Gunicorn worker count on Azure App Service. That got it to 99.9%. That's how I usually work: go past …
Skills
- Amazon S3
- Docker
- JavaScript
- Node.js
- PostgreSQL
- Python
- React
- Redis
- Typescript
- JavaScript ES6
- Amazon AWS
- Amazon SQS
- Flask
- Phyton
- ReactJS
- Python/R
- Python3
- FastAPI
- Celery
- Nuxt.js
- Paython
- SQLAlchomy
- Nest.js
- Next.js
- Next js
- RAG
- Langchain
- Supabase
- Amazon Q
- Fast API
- Type Script
- Apache Airflow
- Pytest
- Claude API
- Typescrtipt
- Amazon ECS
- Javascript
- SQLAlchemy
- Claude AI
- Typescripts
- PostgreQL
- Pyrhon
- TypeScript