FinanceRadar is a small hobby project I built for myself.
There is no grand end goal here.
I just had a practical problem: I was wasting too much time jumping across sources, seeing the same story repeated, and still missing what actually mattered.
So I built a setup that gives me a cleaner research feed.
The problem I was trying to solve
When you track business and market news seriously, the bottleneck is rarely access.
It is signal.
Too much noise, too many repeats, too many low-value updates, and too much context-switching.
I wanted one place that helps me:
- scan faster
- avoid duplicate headlines
- spend more time on analysis than collection
What FinanceRadar does now
Today, it is a lightweight workflow tool that:
- pulls from a broad set of finance/business sources on an hourly cycle
- filters obvious low-signal items and removes duplicates
- groups things in a way that makes scanning easier
- adds a daily AI-ranked shortlist of stories worth paying attention to
- pulls public brokerage-report updates from selected Telegram channels
It is still intentionally simple.
No dashboards for the sake of dashboards. Just a cleaner pipeline.
How I use it
My flow is straightforward:
- Open FinanceRadar and do a quick top-level scan
- Check the AI shortlist for prioritization
- Open only the stories that look worth deeper reading
- Save items I want to come back to while writing
That is it.
The tool exists to reduce friction in this loop.
Current setup (behind the scenes)
The project runs as a static site with automated refresh jobs:
- feed aggregation refreshes hourly
- AI ranking runs daily
- output is published to
financeradar.kashishkapoor.com
So maintenance stays low, and the workflow stays reliable.
If you are curious, you can check it here: