Incerto is an AI-native Copilot for databases. Using context-aware agents, it turns natural language into tasks — like writing queries, schema changes, or moving data — boosting dev productivity 10x with 90% less manual work.
Appreciate your constant support. Yeah idea has evolved quite a bit, many silly observations we missed (like primary job of database is to fetch data out of it not maintain database, we start with DBA's only as primary ICP)
Hope that your team also uses our product in some near future!
Dang, love the “Text-to-Task” approach—making databases actually understand my weird requests without breaking stuff is a game changer. Curious if you’re planning team workflows next?
Yes, doing complex tasks with dynamic context management is what we are working on now. You will see mermaid, and then human loop at each step of the execution
You can build semantic layer over data using training easter eggs.
Selecting a database and typing "train <database_name>" Extracts the information and summary about all the tables.
For certain businesses make the contexts manually too, to get the desired accuracy.
We have recently substituted Tablaue for one business with 100% on the known metrics they track -- because we use the already tested SQLs and tablaue syntax and specialized agent for that.
Feature like system prompt replacement (practically like spawning customer agent), templates, auto templates and DBMS specific templates were exactly meant to create this semantic layer. So that non technical user can just talk to their databases.
If you download the product, it comes with semantic layer generated on dummy dataset.
Hoping for good feedback on the product, thanks for the comment again.
Congrats on the launch, Shiva! Really like the “text-to-task” idea, feels way more practical than just text-to-SQL. What’s been the trickiest part of making the agents actually get the database context right?
Congrats on the launch! Lovely to see how the idea has developed!
@raja_mahapatra thanks bhai!
Appreciate your constant support. Yeah idea has evolved quite a bit, many silly observations we missed (like primary job of database is to fetch data out of it not maintain database, we start with DBA's only as primary ICP)
Hope that your team also uses our product in some near future!
Incerto
@raja_mahapatra bhai, Thank you ❤️ for all the constant support.
bas aise hi support karte raheye, we will reach somewhere.
Shiva pandey all the best .. So proud of you guys.. Really done a great job.
Incerto
Shifu, Thank you ❤️@rajib_debbarma1
(─ ‿ ─)
Agnes AI
Dang, love the “Text-to-Task” approach—making databases actually understand my weird requests without breaking stuff is a game changer. Curious if you’re planning team workflows next?
@cruise_chen thanks for the kind words!
Once can make templates to make some workflow.
Yes, doing complex tasks with dynamic context management is what we are working on now. You will see mermaid, and then human loop at each step of the execution
It's actually part of current sprint :)
Looking forward to your feedback!
@cruise_chen I misunderstood.
Yes, team collaboration is still not supported! We will work on it
This looks amazing! 🎉
Can Incerto answer questions by combining what’s in the database with the business rules/semantics on top of it?
@aishwarya_prakash2 thanks for asking it, yes!
You can build semantic layer over data using training easter eggs.
Selecting a database and typing "train <database_name>" Extracts the information and summary about all the tables.
For certain businesses make the contexts manually too, to get the desired accuracy.
We have recently substituted Tablaue for one business with 100% on the known metrics they track -- because we use the already tested SQLs and tablaue syntax and specialized agent for that.
Feature like system prompt replacement (practically like spawning customer agent), templates, auto templates and DBMS specific templates were exactly meant to create this semantic layer. So that non technical user can just talk to their databases.
If you download the product, it comes with semantic layer generated on dummy dataset.
Hoping for good feedback on the product, thanks for the comment again.
A01
Congrats on the launch, Shiva! Really like the “text-to-task” idea, feels way more practical than just text-to-SQL. What’s been the trickiest part of making the agents actually get the database context right?
Incerto
Thank you ❤️ @shengkun_ye
I would say context and memory management. But Anurag would be the right person to do in details :)
Astra Security
Feels like magic 🪄
@anandakrishna thanks <3
Incerto
@anandakrishna I hope so. haha.
Thank you for the support ❤️
Wish you the best Shiva and Anurag!
@pranav_khambayatkar Thanks Pranav Ji, thank you for you constant support.
Incerto
@pranav_khambayatkar
Pranav sir, thank you very much ❤️. We will continue to work harder.