It's an automation tool for creating widgets. In DevRev, our Support and Applied AI teams create lots of dashboards for customers as well as for Dev0. These dashboards consist of multiple components - Widgets.
I thought, why not automate this? We're doing the same repetitive work regularly. The widget creation process follows the same steps every time - analyze data, write SQL queries, configure widget JSON, test. So why not build an AI automation that does all of this for us?
That's the reason for me to create it.
Why??
We need it regularly - Support and AAI teams create widgets daily for customer requests and internal dashboards
We're doing it manually right now - Even with tools like Cursor, it's still a lot of manual work
It takes too much time - Writing SQL queries, configuring widget JSON, testing - easily 1-3 hours per widget
If automated → We can spend that time on more valuable work, or honestly, just have some free time 😊
Working flow
1
Upload CSV Files
Upload your CSV file(s) - the tool can handle up to 3 CSV files simultaneously for multi-dataset analysis.
2
AI Analysis & Query Suggestions
AI automatically analyzes the dataset structure and generates suggested SQL queries based on the data patterns it detects.
3
Custom Query Generation (Optional)
Use the AI search feature to generate custom queries beyond the suggestions. Just describe what you need in natural language.
4
Select Query & Chart Type
Choose your preferred query from the suggestions or custom queries, and select the visualization type (table, line, column, pie, etc.).
5
Generate Widget JSON
The tool automatically generates the complete widget JSON configuration that's ready to use in DevRev.
6
Copy & Paste
Copy the generated widget JSON and paste it directly into DevRev. Your widget is ready to use.
See It In Action
How's This Different?
You might ask: "Why do we need this? We already have Cursor, right?"
That's a fair question! Here's how this is different:
Multiple dataset analysis - Cursor can't analyze multiple datasets at once. Our tool can handle 3 CSV files simultaneously. It generates SQL queries for each dataset individually AND automatically creates JOIN queries when it detects relationships between tables. Cursor helps with coding, but it doesn't automate the entire query generation workflow.
Complete end-to-end automation - With Cursor, you still need to:
Understand your data structure
Write SQL queries manually
Create widget JSON configurations from scratch
Test and debug everything
Our tool does ALL of this automatically. Just upload your CSV files → Get SQL queries → Get widget JSON → Copy and use. That's it.
The Impact
Current State:
Developer writes SQL query manually (30 mins - 2 hours depending on complexity)
Developer creates widget JSON configuration (30 mins - 1 hour)
Testing and debugging (15-30 mins)
Total: 1-3 hours per widget
With Automation:
Upload CSV → Get queries → Get widget JSON → Copy-paste → Done
Total: 2-5 minutes per widget
That's 10x faster! ⚡
Pros & Cons
✅ Pros
Saves hours of developer time every single day
Support and AAI teams can work independently without blocking developers
Users need basic understanding of their data structure (but no SQL knowledge needed)
Real Example
Before
Support team needs: "Show me all tickets by status for customer X"
They try to create it manually or wait for developer
Takes 1-2 hours to complete
Often has errors and needs fixing
Result: Slow turnaround, potential errors, time wasted
After
Support team needs: "Show me all tickets by status for customer X"
Uploads CSV file
Gets query in 30 seconds
Gets widget JSON in 1 minute
Result: 2 minutes total, works perfectly, no developer needed
Why Push to Live?
One "yes" can change everything - the workload, bandwidth, and productivity.
High demand - Teams need this functionality daily
Immediate impact - Reduces developer workload from day one
Low risk - Works seamlessly with existing DevRev infrastructure
Scalable - Handles any data structure automatically
User-friendly - Non-technical users can adopt it quickly
Success Metrics:
80% reduction in query generation time
50% reduction in developer hours spent on routine queries
70%+ self-service rate for support/AAI teams
90% fewer widget configuration errors
Future Vision
We already have a computer application. Why can't we just tell it what we need?
$ widget create
"Using these datasets, create a widget showing ticket status by customer"
✓ Widget created successfully
Today
Upload CSV→AI suggests
Select query→Generate JSON
Copy & paste
Tomorrow
"Show me ticket status by customer"
↓
Widget created automatically
When customers can build what they need instantly, without barriers → Customer adoption increases → Growth follows.
Bottom Line
This is not just a tool - it's a productivity multiplier.
We're spending hours every day doing the same repetitive work. Why not automate it? The time saved can be spent on more valuable work, or honestly, just having some free time 😊
Looking forward to your positive response to take it to the next level from version 0.