https://github.com/ragul-kachiappan/hearth
Simple Chat room app with LLM support
https://github.com/ragul-kachiappan/hearth
async aws chromadb docker gcp generative-ai k8s mlops python redis starlette terraform websocket
Last synced: 3 months ago
JSON representation
Simple Chat room app with LLM support
- Host: GitHub
- URL: https://github.com/ragul-kachiappan/hearth
- Owner: ragul-kachiappan
- License: mit
- Created: 2024-12-25T07:57:28.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-02-08T14:22:17.000Z (over 1 year ago)
- Last Synced: 2025-02-08T15:26:30.734Z (over 1 year ago)
- Topics: async, aws, chromadb, docker, gcp, generative-ai, k8s, mlops, python, redis, starlette, terraform, websocket
- Language: Python
- Homepage:
- Size: 70.3 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# hearth
`STATUS: Side Project on hold. Will resume when learning js`
Tech Stack
- Django for API server
- Websockets for socket server
- Postgres (could be NoSQL)
- std html, css, js for frontend. With htmx and lightweight frameworks like alpine. Full react frontend for next phase
- Containerized with Docker
- Deployed with k8s along with ArgoCD. (microk8s or something similar)
Specialities about the project.
I'm building it mainly to test my newly acquired skills and deep dive into it.
- End to End project with main focus on backend concepts.
- Fully async server including the DB client.
- try out uv for package management.
- websocket support.
- Redis for caching.
- Background workers.
- Starlette, Redis, ChromaDB, Postgres, Nginx container setup.
- Logging to sentry.
- Possibly Oauth with gmail. Not mandatory.
- Proper test suite.
- Proper markdown docs.
- Deployment to GKE with GCP free credits
- cicd for linting checks (ruff, yaml, toml, js, html, css).
- prometheus, elk stack for monitoring and logging.
- perform server profiling with wrk (sync vs async)
- NLP Abuse detection
- maybe terraform for provisioning some AWS resources like ECR, Bucket, SSM
Road to k8s deployment
[x] - Docker
[ ] - Docker optimization
[ ] - local k8s setup
[ ] - terraform for AWS resources
[ ] - helm chart
[ ] - manual deployment
[ ] - Argo, csi secrets store, istio setup
[ ] - cicd for docker build and push, argo sync
[ ] - GKE setup
[ ] - Deepseek R1/LLama for local server, Gemini free tier for cloud
I want this project to encapsulate everything I have learnt regarding backend, devops and AI.
```
Environment Variables:
Should use more secure password management in production
Logging:
Could add log rotation
Might want to consider adding a logging service like ELK stack
Monitoring:
Could add Prometheus/Grafana for metrics
Health check endpoints could be more comprehensive
Scaling:
Could add replicas for the web service
Might want to consider using Docker Swarm or Kubernetes for larger deployments
```