Streamlit Real-Time Ops Dashboard -- 2
Budget: $250 – $350 USD
I need an internal control panel that focuses on real-time data visualization. The stack is fixed—Streamlit for the UI, Redis as the live data feed, and Postgres for longer-term storage—so you should already be comfortable wiring these pieces together.
Here’s what I’m after:
• A Streamlit app that auto-refreshes with live metrics pulled from Redis while persisting snapshots to Postgres for later queries.
• Clear, attractive charts and widgets that let the team filter, sort and drill down without page reloads. Use Plotly, Altair or any Streamlit-friendly library you prefer, as long as the visuals remain smooth at high update rates.
• A lightweight auth layer (Streamlit’s built-in, or something similarly simple) so only our ops crew can see the dashboard.
• Container-ready code: Dockerfile, requirements.txt and a concise README so I can spin it up with a single command on our internal server.
• Unit tests for the key data pipelines—Redis listener, Postgres writer, and any helper utilities—to make sure nothing breaks when we tweak things later.
Acceptance criteria
1. Live metrics appear on screen within one second of being pushed into Redis.
2. Historical views load from Postgres in under two seconds for the past 24 h range.
3. All charts render correctly on Chrome, Firefox and Edge at 1080p.
4. The entire stack starts with `docker compose up` and shows no errors in the logs.
If any details on the specific metrics or chart types are still unclear, flag them early and I’ll clarify. Looking forward to seeing how you bring real-time visibility to our operations.
Here’s what I’m after:
• A Streamlit app that auto-refreshes with live metrics pulled from Redis while persisting snapshots to Postgres for later queries.
• Clear, attractive charts and widgets that let the team filter, sort and drill down without page reloads. Use Plotly, Altair or any Streamlit-friendly library you prefer, as long as the visuals remain smooth at high update rates.
• A lightweight auth layer (Streamlit’s built-in, or something similarly simple) so only our ops crew can see the dashboard.
• Container-ready code: Dockerfile, requirements.txt and a concise README so I can spin it up with a single command on our internal server.
• Unit tests for the key data pipelines—Redis listener, Postgres writer, and any helper utilities—to make sure nothing breaks when we tweak things later.
Acceptance criteria
1. Live metrics appear on screen within one second of being pushed into Redis.
2. Historical views load from Postgres in under two seconds for the past 24 h range.
3. All charts render correctly on Chrome, Firefox and Edge at 1080p.
4. The entire stack starts with `docker compose up` and shows no errors in the logs.
If any details on the specific metrics or chart types are still unclear, flag them early and I’ll clarify. Looking forward to seeing how you bring real-time visibility to our operations.